advanced gas tech i toc+preface - home | official web ...€¦ · content preface vii chapter 1...

First Edition 2006 © KHAIRUL SOZANA NOR KAMARUDIN & MARIANI IDROAS 2006

Hak cipta terpelihara. Tiada dibenarkan mengeluar ulang mana-mana bahagian artikel, ilustrasi, dan isi kandungan buku ini dalam apa juga bentuk dan cara apa jua sama ada dengan cara elektronik, fotokopi, mekanik, atau cara lain sebelum mendapat izin bertulis daripada Timbalan Naib Canselor (Penyelidikan dan Inovasi), Universiti Teknologi Malaysia, 81310 Skudai, Johor Darul Ta’zim, Malaysia. Perundingan tertakluk kepada perkiraan royalti atau honorarium. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopy, recording, or any information storage and retrieval system, without permission in writing from Universiti Teknologi Malaysia, 81310 Skudai, Johor Darul Ta’zim, Malaysia. Perpustakaan Negara Malaysia Cataloguing-in-Publication Data Advanced gas technology. 1 / editors Khairul Sozana Nor Kamarudin, Mariani Idroas. ISBN 978-983-52-0598-9 1. Gas engineering. 2. Gases. I. Khairul Sozana Nor Kamarudin, 1965-. II. Mariani Idroas. 665.7

Editor: Khairul Sozana Nor Kamarudin & Rakan Pereka Kulit: Mohd Nazir Md. Basri & Mohd Asmawidin Bidin

Diatur huruf oleh / Typeset by Fakulti Kejuruteraan Kimia & Kejuruteraan Sumber Asli

Diterbitkan di Malaysia oleh / Published in Malaysia by PENERBIT

UNIVERSITI TEKNOLOGI MALAYSIA 34 – 38, Jln. Kebudayaan 1,Taman Universiti

81300 Skudai, Johor Darul Ta’zim, MALAYSIA.

(PENERBIT UTM anggota PERSATUAN PENERBIT BUKU MALAYSIA/ MALAYSIAN BOOK PUBLISHERS ASSOCIATION dengan no. keahlian 9101)

Dicetak di Malaysia oleh / Printed in Malaysia by

UNIVISION PRESS SDN. BHD Lot. 47 & 48, Jalan SR 1/9, Seksyen 9,

Jalan Serdang Raya, Taman Serdang Raya, 43300 Seri Kembangan,

Selangor Darul Ehsan, MALAYSIA.

CONTENT

Preface vii

Chapter 1 Computational Fluid Dynamics Modelling In Natural Gas Orifice Metering Station With Straightening Vane 1 Natasha Noorazrin Bt Mohd Basir Zulkefli Yaacob, Norhaniza Yusof

Chapter 2 A Study On Air-Fuel Mixing System For Ngv Motorcycle System 31 Zulkifli Abdul Majid, Rahmat Mohsin, Nor Faizal Bin Nordin

Chapter 3 Finite Element Structural Analysis On A Single Step NGVM Pressure Regulator BodyRahmat Mohsin, Khairul Sozana Nor 59

Chapter 4 Emission And Performance Evaluation Of Motorcycle 61 Zulkifli Abdul Majid, Zulkefli Yaacob, Rahmat Mohsin

Chapter 5 Water Jetting Impact On Natural Gas Piping (Abrasion Effect And Metal Thinning) 97 Zulkifli Abd Majid, Rahmat Mohsin, Fadhli Omar

Chapter 6 Vapor Liquid Equilibrium Behaviour Of Liquefied Petroleum Gas In Storage 115 Zainal Zakaria, S.Y. Tee

Chapter 7 Analysis Of Household Energy Demand 129 Zainal Zakaria, Wong Song Harn

Chapter 8 Newton Loop Method In Gas Pipeline Network 145

Zulkefli Yaacob, Norhana Mohamed Rashid

Index 175

vi

PREFACE

First, I thanked Allah Almighty for His guidance and strength, who has helped us to complete a series of Book Chapters titled Advanced Gas Technology I.

Advanced Gas Technology I is a compilation of researches done at the Department of Gas Engineering, Faculty of Chemical and Natural Resources Engineering (FKKKSA). The topics covered in this book are related to recent technologies in gas engineering and/or gas industry. I hope this book will be useful to engineers, postgraduate students, academician and anybody interested in the development of gas technology.

Many people have assisted during the preparation of this book and I would like to thank them for the comments and suggestions they have made.

Khairul Sozana Nor Kamarudin Mariani Idroas Universiti Teknologi Malaysia 2006

1COMPUTATIONAL FLUID DYNAMICS

MODELLING IN NATURAL GAS ORIFICE METERING STATION WITH

STRAIGHTENING VANE Natasha Noorazrin Bt Mohd Basir

Zulkefli YaacobNorhaniza Yusof

INTRODUCTION

In the gas industry, a reliable, accurate and efficient measurement of gas quantities is of the utmost importance. In fact, an accurate metering of gas flow rates assures the rightness of the commercial transactions (which, of course, have economic implications) and the improvement of pipeline efficiency, by means of telecontrol systems (Cascetta and Scalabrini,1998). A flowmeter for the natural gas industry must have some specific technical characteristics. In practice, an ideal gas flowmeter should be (Cascetta and Scalabrini,1998):

• Accurate and repeatable, according to the application (custody transfer metering or measurement in telecontrol);

• Static (i.e. without moving parts) in order to reduce the maintenance costs and to have long-term stability and repeatability;

• Safe, in order to reduce risks in hazardous area; • Obstruction less, in order to allow the transit of a robotized

system (mouse, pig), normally used for the inner inspection/cleaning of the pipeline.

2 Advanced Gas Technology I

• Non-intrusive, in order to reduce pressure drops; • Flexible, in order to satisfy a wide flow range

THE ORIFICE METERING

The Orifice Plate

The most common differential-pressure type flow meter used in pipelines is the sharp-edged orifice plate. Applications with proper water quality, careful attention to installation detail, and proper operation techniques (Hobbs, 1987) make these flow meters capable of producing accuracy to within 1 percent. However, the usual maintenance and pipe conditions that generally occur in irrigation pipe systems limit field accuracies to within 3 to 5 percent of actual. The orifice plate is commonly used in clean liquid, gas, and steam service. It is available for all pipe sizes, and if the pressure drop it requires is free, it is very cost-effective for measuring flows in larger pipes (over 6" diameter). The orifice plate is also approved by many standards organizations for the custody transfer of liquids and gases (Instrument Society of America, 1989). Based on the AGA Report No.3 (1985), the primary consideration in the design of a metering station is to sustain accuracy. Many particulars must be considered and assembled into specifications and drawing of the meter tube and its installation in the piping system to provide the accuracy, safety and also work space for operations. According to the metering station that is being considered in this study, the author refers to the NPS 6 pipes. Using Figure 2.1 with Table 2.1, the recommended orifice plate thickness is 0.125 inches. The nominal inside diameter of the orifice plate, d is 4.897 until 6.065 inches. The maximum orifice

Computational Fluid Dynamics Modelling 3

edge thicknesses is defined by e D/50 or e d/8. So, the value of e is equal to 1/16 inches until 7/64 inches; depends on the orifice diameter chose. This parameter has been selected because it has satisfied the orifice plate dimension stated in AGA Report No. 3 and avoiding uncertainty due to exceeding the required limits.

Figure 1.1 Orifice Plate Dimension (AGA Report No.3)


Table 1.1 Orifice Plate Dimension (AGA Report No 3)

The upstream face of the orifice plate shall be perpendicular to the axis of the meter tube. The departure from flatness shall be determined in Figure 2.2. Based on the nominal meter tube size that is equal to 6 inches, the maximum departure from flatness is equal to 0.008 inches. The mean diameter orifice is defined as the arithmetic average of four or more inside diameter measurements evenly spaced. It should be used in the coefficient equations for the calculation of the flow coefficient with minimum uncertainty. The mean diameter should not differ more from the diameter used in a simulation. Practical tolerances for orifice diameters are shown in Table 2.2. The orifice must be places concentric with the inside of meter tube. It shall retained within 3 percent of the inside diameter of both upstream and downstream section of the orifice meter tube.


Table 1.2 Practical Tolerances for Orifice Diameters (AGA Report No.3)

Orifice Diameter, d Tolerance, plus or minus

0.250 0.0003 0.375 0.0004 0.500 0.0005 0.625 0.0005 0.750 0.0005 0.875 0.0005 1.000 0.0005 Over 1.000 0.0005 per inch of diameter

There are several advantages and disadvantages of the orifice plate. The advantages of the orifice plate are its simplicity and the ability to select a proper calibration on the basis of the measurements of the geometry. Disadvantages of the orifice plate include the long, straight pipe length requirements and the limited practical discharge range ratio of about one to three for a single orifice hole size.

Meter Tube

Meter tube is a combination of a straight upstream pipe of the same diameter (of length A or A’, including the straightening vanes if used), the orifice flanges or fittings, and the similar downstream pipe (length B) beyond the orifice.


Figure 1.2 Less than Ten Pipe Diameter (D) between Two Ells in Same Plane Upstream of Meter Tube (AGA Report No.3)

Normal flow conditions were obtained by the use of straight lengths of meter tube, both upstream and downstream from the orifice. Any distortion of the flow profile will cause errors. There are some restrictions and tolerances for these particular cases for the inside surface of the meter tube such as: 1. The difference between the maximum and minimum measured

diameter on the inlet section shall not exceed the tolerance allowed by figure 1.3. The equation to calculate the variance of the upstream section of the meter tube is as bellow (AGA Report No.3):

MaxDiamete r MinDiamete r 100 % tolerance in Figure 2.3 (1.1) D

2. Abrupt changes in diameter shall not exist in meter tube.


3. Any diameter measurement in the downstream section shall not vary from the mean diameter of the meter tube.

4. The temperature at which the meter tube measurements are made should be recorded for possible correction to operating conditions.

Figure 1.3 Maximum Percent Allowable Meter Tube Tolerance

The Straightening Vanes

Straightening Vanes are installed in the upstream section of meter tubes to reduce flow disturbance preceding the orifice plate. Disturbance is often created by complex piping or valves which precede the orifice metering section. As flow passes through the vane bundle, the disturbance is straightened and smoothed to a normal flow pattern. Straightening Vanes are economical because their use often allows sufficient reduction of upstream meter tube


length, so that a smaller building or enclosure is possible. The tubes in each vane bundle are welded at both ends at each point of tangency. Tube inlets and outlets are reamed to permit minimum pressure drop. Special spacer lugs on each vane bundle assure a perfect fit in the meter tube.

Figure 1.4 Standard Straightening Vane Bundles (cross-section end view)

In order to design a bundle of vanes, the maximum inside diameter of tubes shall not exceed one-fourth (1/4) the line inside diameter, D. Also, the cross sectional area, A, of any passage within the assembled vanes shall not exceed one-sixteenth (1/16) of the cross sectional area of the containing meter tube. The length, L of the vanes shall be at least ten times the maximum inside dimension, a (AGA Report No.3).


Figure 1.5 Straightening Vanes


Figure 1.6 Line Model of Carbon Steel Straightening Vanes

Table 1.3 Line Model Straightening Vane (Daniel, 1989)


Reynolds Number

Reynolds number is the ratio of inertial forces (vs ) to viscous forces ( /L) and consequently it quantifies the relative importance of these two types of forces for given flow conditions. Thus, it is used to identify different flow regimes, such as laminar or turbulent flow.

The Turbulent Kinetic Energy, k and Rate of Dissipation,

Turbulent kinetic energy refers to the portion of the total kinetics energy of the fluid that is presented in the velocity fluctuations. Viscous dissipation presented laminar flow and it’s proportional to the fluid viscosity and the square root of the velocity gradient. The dissipation rate of turbulence is the rate at which turbulent kinetic energy is converted o the internal energy of the fluid by molecular viscosity. The dissipation is equals to the rate of change of kinetic energy in a fully developed flow. The turbulent contribution to the stress tensor is called the Reynolds’ stress. The motion of fluid is governed by Newton’s 2nd law of motion in the form of the Navier-Stokes equation (Biswas and Eswaran, 2002),

METHODOLOGY

Selection of Metering Station

First, the metering station is being studied and the part of meter tube is measured. The diameter of meter tube, the thickness of the orifice plate, the length of the straight pipe and the temperature of


gas flow has been measured. Then, the diameter of the tube, 6 inches is fixed. The meter tube is fixed because one of utility company in Malaysia used meter tube with 6 inches in their metering station. Therefore it became guidance in this study and measurement of orifice thickness and orifice bore diameter were depends on meter tube selected.

Selection of Beta Ratio

In this study, the orifice bore diameters selected were 0.625, 0.75 and 0.875 inches because the author now compares two different sizes for easier comparison. These diameters were chosen from table 2.1, with the recommended nominal inside diameter for 6 inches meter tube is 6.0625 inches and the orifice plate thickness is 0.125 inches. Table 1.1 also indicates maximum orifice edge thickness, e. Thus, the ß ratio in this study is 0.104, 0.125 and 0.146. The beta ratio selected may have uncertainty flows about ±1 % due to smaller ß ratio compared to approximate one, which is 0.20 to 0.70.

The Selection of Meter Tube Length

After beta ratio is calculated, select the length of meter tube. There are only one case that were taken into consideration in this analysis, which are 5D (in the same plane & separated by less than ten diameters of straight pipe) preceding the straight pipe of meter tube. 5D cases were chosen owing to the various types of installation the meter tube. Therefore, it is important to select the right figures in AGA No. 3 for easier model creation in GAMBIT.


Furthermore, one can compared the best length of meter tube chose and determined which figures gave the finest result.

The Selection of Straightening Vane Parameters

In this study, carbon steel straightening vane will be select. Since the meter tube is 6 inches, straightening vanes with 12 inches length of vane and 19 numbers of tubes was chosen. The recommended outside diameter of vane and tubes are 5 15/16 and 13/16.

Model Creation

By using GAMBIT 2.3.0.0, the model of metering station is build. Based on the length of meter tube, orifice bore diameter (orifice plate) and bends before and after the meter tube; the geometry of orifice meter with straightening vane is creating.

Execution in FLUENT

Before the model is execute in FLUENT, the model must initially mesh, check the grid and specify the boundaries and correct solver is chosen. One must check the grid to ensure there are no missing faces or element during execution in FLUENT. Then, the boundaries condition is specify to obtain correct solution. In FLUENT 6.1.22, the author needs to state the model equation so the computer can solve the problem. Some iteration is essential to make sure the k and are converged. The overall steps are summarized in Figure 1.6


Figure 1.6 Overall steps in modeling of orifice meter


SIMULATIONS

Computational Fluid modeling (CFD) refers to the art of replacing the governing partial differential equations of fluid flow with numbers and advancing these numbers in space or time to obtain a final numerical description of a complete flow field of interest. CFD solutions generally required the repetitive manipulation of millions of numbers that is impossible for human to solve it without the aid of computer (Wendt, 1991).


Figure 1.7 The overall procedure in CFD modeling

RESULTS AND DISCUSSION

In this analysis, there were two cases that have been executed by using FLUENT. The first case is an orifice meter without


straightening vane (Table 1.4) and the other one is an orifice meter with straightening vane installations (1.5). Both cases were based on the meter tube installation with two bends preceding the meter tube and separated by less than 10D of the straight pipe. There are different value of beta ratio, ß that were analyzed for both cases which is 0.104, 0.125 and 0.146. Both cases were divided into four different length of meter tube, namely 1/2 of Standard (Std), 3/4 of Std, 1Std and 5/4 of Std. According to AGA Report No.3, standard length of A and B for meter tube without straightening vane are:

Table 1.4 Standard Length of Meter Tube without Straightening Vane

Beta Ratio, ß A, in B, in 0.104 84 15 0.125 85.2 16.2 0.146 8.4 16.8

On the other hand, there were four different lengths that need to be considered for the installation of straightening vane in orifice meter. The standard length for this case is presented in Table 1.5.


Table 1.5 Standard Length of Meter Tube for Orifice Meter with Straightening Vane

Beta Ratio, ß A, in B, in C, in D, in 0.104 57 15 30 27.6 0.125 57 16.2 30 27.6 0.146 57 16.8 30 27.6

The parameters that are involved in this analysis are Reynolds Number (Re), Turbulent Kinetic Energy (K) and Dissipation Rate ( ). By analyzing the contour at the end of length B, the parameters can be easily determined.

Figure 1.8 The Contour Region

Turbulent Kinetic Energy, k

In this section, the k values for both cases attained from the simulation were compared to the true value. Turbulent kinetic energy, k refers to the portion of the total kinetics energy of the fluid that is presented in the velocity fluctuations. The true value of


k is 5.489 x 10-2 ft2/s2. (Biswas and Eswaran, 2002) Results for both cases at 0.104 beta ratios were presented in Table 1.6.

According to the above results, only one case exceeds the reference value. At 3/4 of Std, orifice meter without straitening vane have higher K value than 5.489 x 10-2 ft2 /s2 (Biswas and Eswaran, 2002). Higher flow rates and velocity after passing through the orifice were disrupted by eddies moving in all flow directions. This makes the flow turbulent and increased the turbulent kinetic energy.

Table 1.6 Turbulent Kinetics Energy at 0.104 Beta Ratios

K ( ft2/s2)Length of B (ß=0.104) Orifice meter without

straightening vane Orifice meter with straightening vane

½ Std 7.5 in 1.93 x 10 -2 2.17 x 10-2

¾ Std 11.25 in 7.92 x 10-2 2.75 x 10-2

1 Std 15 in 1.19 x 10-2 1.79 x 10-2

5/4 Std

18.75 in 0.91 x 10-2 2.61 x 10-2

Based on Figure 1.9, it can be seen that at the end of 5 /4 Std length of B produced the best contour and with the lowest K value. However, with straightening vane installation, the length of B can be shorted to 3/4 Std (Figure 1.9). Consequently, the same parameter is also analyzed for 0.125 beta ratio. As presented in Table 1.7, it can be seen that all value are lower than the references. This is because the total kinetic energy of the fluid was proportional to the intensity of turbulence. It shows that at this beta


ratio, the flow is least turbulence and has the tendency become laminar.

Figure 1.9 Turbulent Kinetics Energy for Orifice Meter without Straightening Vane on Beta Ratio 0.104

Figure 1.10 Turbulent Kinetics Energy for Orifice Meter with Straightening Vane on Beta Ratio 0.104





½ Std 8.1 in 3.01x 10 -2 2.14 x 10-2

¾ Std 12.15 in 2.93 x 10-2 2.93 x 10-2

1 Std 16.2in 3.27 x 10-2 2.60 x 10-2

5/4Std

20.25 in 3.65 x 10-2 2.59 x 10-2

Finally, an analysis for turbulence kinetic energy was done for 0.146 beta ratios. As recorded in Table 1.8, it can be seen that all cases have lower K value when compared to standard value which is 5.489 x 10-2 ft2/s2 . (Biswas and Eswaran, 2002). These results shows that the flow in pipe for every case is less turbulence at 0.146 beta ratio.





½ Std 8.4 in 3.01x 10 -2 2.14 x 10-2

¾ Std 12.6 in 2.93 x 10-2 2.35x 10-2

1 Std 16.8 in 3.27 x 10-2 2.60 x 10-2

5/4 Std

21 in 3.65 x 10-2 2.59 x 10-2

It is suggested that, both orifice meter with and without straightening vane can be shorted to 1/2 Std. However, this results need more consideration on others parameter such as dissipation rate and Reynolds number.

Rate of Dissipation,

The values of dissipation rate, attained from the simulation are presented in this section. Rate of dissipation, refers to the rate at which K is converted to the internal energy of the fluid by molecular viscosity. (Biswas and Eswaran, 2002) mentioned that the true value of dissipation rate, is equal to 3.154 x 10-2 ft2/s3.This standard value has been compared to the simulation results.


As in previous analysis, this parameter has been analyzed based on three different beta ratios which are 0.104, 0.125 and 0.146. Table 1.9 shows the rate of dissipation at four different lengths for 0.104 beta ratio. According to the simulation results, installation of orifice meter is more suitable without straightening vane at lower beta ratio. As shown in Table 5.6, orifice installations with straightening vane give higher value compared to the standard. In meter tube, relatively large rotational eddies form in regions of high shear near the pipe wall. These degenerate into smaller eddies as energy is dissipated by action of viscosity. Therefore, when gas velocities increased in meter tube with low viscosity, it tends to dissipate the flow.

Table 1.9 Dissipation Rate, at 0.104 Beta Ratios

The dissipation rate of the beta ratio 0.125 is shown in Table 1.10. All the value is higher than the standard value. It shows that the values are close to the theoretical and can be considered in the pipe length selection. Although, 5/4 Std length of B show the best result among the four different length, other cases can be


considered based on the another two parameters, that is Reynolds number and K values.


For beta ratio of 0.146 at four different lengths, both cases have higher dissipation rate than the reference value respectively (Table 1.11). However, these results are still close to the standard.



Reynolds Number, Re

The last parameter discussed in this section is the Reynolds Number, Re. This parameter describes the relative importance of molecular and convective transport in a flowing stream. Since molecular transport dominates in laminar flow and convective transport in turbulent flow, the Reynolds number also serves as an indicator of the flow regime (AGA Report No.3). Figure 1.11 shows the simulation results of the Re for both cases. The maximum value of Reynolds number at certain distances was recorded. For each beta ratio, comparisons between both cases were done by plotting a graph of Reynolds number versus distance. There are different limitations on Re values for every case. Meter tube without straightening vane has bigger limitation, that is from 50 to 4375000, than meter tube with straightening vane which the limitation is from 50 to1708000. Conventionally, be laminar, the flow should have a Re value of 2000 or less. Based on AGA Report No.3, length of B should be 15 inches at beta ratio equal to 0.104. However, after observation on the results, it is suggested that at beta ratio 0.104, length of B should be longer than the


standard, which is 18 inches for orifice meter without straightening vane. While the length of B for orifice meter with straightening vane should be extended to 16 inches. This is because, in both cases, the Reynolds number is lower than 2000 at this length.

Figure 1.11 Reynolds Number of Orifice Meter with Straightening Vane on Beta Ratio 0.104

Figure 1.12 shows the comparison between meter tube with and without straightening vane at 0.125 beta ratio. As shown in the graph, outlet flow from the orifice was at low Reynolds number but not enough to be laminar. In the intermediate the flow was at higher turbulence occurs at high Re and is dominated by inertial


forces, producing random eddies, vortices and other flow fluctuation. After that, the flow starts to smooth and with lesser disturbance and thus becomes laminar at the end of length B. Although AGA Report No.3 suggested that the length of B at this beta ratio equal to 16.2, but these results proved that 14.5 inches of length B is enough for the installation of an orifice meter without straightening vane. On the other hand, installations orifice meter with straightening vane required 14 inches of length B, which is shorter than the standard value.

Figure 1.12 Reynolds Number on Beta Ratio 0.125


As in previous section, the last beta ratio for the discussion is 0.146. Simulation results for this beta ratio have been presented in Figure 1.13. As shown in the graph, meter tube with straightening vane has the smallest range of Re. This demonstrates that the installation of a straightening vane in meter tube can reduce the turbulent and also cause the flow to become smooth and with lesser disturbance.

Figure 1.13 Reynolds Number on Beta Ratio 0.146

Besides, the simulation results demonstrated that both cases have shorter length of B compared to the standard from AGA Report No.3. As shown in the graph below, meter tube with


straightening vane can be reduced to 10 inches. While without straightening vane installations the meter tube can be reduced to 14 inches. This result produced much shorter length as compared to the standard, which suggested that the length of B is equal to 16.8 inches. This is because at this length, the Re is below the standard value. This analysis shows that the suggested length of B from AGA Report No.3 can be adjusted to be shorter than the standard length that is also agree with Kelner and Morrow (1999).

CONCLUSION

Study concluded that the AGA standard for length B in orifice meter installation with and without straightening vane can be shortened. For beta ratio 0.125, the length was shortened from 16.2 inches to 14.5 inches without straightening vane and 14 inches with straightening vane. Meanwhile, for beta ratio 0.146, the length was shortened from 16.8 inches to 14 inches without straightening vane and 10 inches with straightening vane. However, for beta ratio 0.104, the length B was lengthen from 15 inches to 18 inches without straightening vane and 16 inches with straightening vane. The author proved that the pipe length for orifice meter with straightening vane is shorter than pipe without straightening vane as suggested by AGA Report No.3. The optimum length of B without straightening vane for beta ratio 0.104, 0.125 and 0.146 are 18, 14.5 and 14 inches respectively. The optimum length of B with straightening vane for beta ratio 0.104, 0.125 and 0.146 are 16, 14 and 10 inches respectively.


REFERENCES

Cascetta F., G. Scalabrini, (1998). Field Test of a Swirlmeter for Gas Flow Measurement. Italy: DETEC, University of Naples.

A.G.A Report No.3,(1985). Orifice Metering of Natural Gas.Arlington: American Gas Association. 13-18

FLUENT, (1999).Gambit Tutorial Guide. Lebanon, New Hampshire: Fluent Incorporate. 0-6

Biswas, G., V. Eswaran, (2002). Turbulence Flows: Fundamentals, Experiments and Modelling. Kampur, India: Alpha Science International Ltd. 1-11

Kelner, E., Morrow. T.B. (1999). “Compact Orifice Meter Station Project”. 4th International Symposium on Fluid Flow Measurement. Denver, Colorado: 1-14.

2A STUDY ON AIR-FUEL MIXING

SYSTEM FOR NGV MOTORCYCLE SYSTEM

Zulkifli Abdul Majid Rahmat Mohsin

Nor Faizal Bin Nordin

INTRODUCTION

The way natural gas engine system works is slightly different than the engine system that used gasoline as it raw fuel material. Basically in the natural gas engine system, natural gas is compressed to 3000 psi and stored on board the vehicle in cylinders (Zulkifli et al., 1999). When natural gas powers the engine, it leaves the storage cylinders, passes through a master manual shut-off valve and travels through stainless steel lines to a high pressure fuel regulators located in the engine compartment. Then the natural gas will be injected at atmospheric pressure through a specially designed gas mixer where it is properly mixed with air. The air-fuel mixture then flows into the engine’s combustion chamber where it is ignited to create the power required to drive the vehicle. Special solenoid valves will prevent the gas from entering the engine when it is shut off.

Gas mixer is a device used to determine the amount of natural gas mixed with air before entering the engine and capable of providing the engine with the appropriate mixture of fuel and air according to the different power regimes and engine revolutions (Rosli et al., 2001; Akira et al., 1999). The perfect mixing of


natural gas and air basically will help to increase the performance of the NGV engine and produced much cleaner exhaust emissions.

Gas mixer is believed to be capable to give the right air-fuel ratio before letting the mixed air-fuel flows into the engine’s combustion chamber. It is also reported that high performance of a NGV engine can be achieved when a good combustion occurred in the power stroke of the engine (James and Rinchard, 1979). However to achieve a good combustion, a correct air-fuel mixture is required based on the air-fuel ratio. The correct air-fuel mixture provides enough air to completely burn all the natural gas (fuel) injected in the engine’s combustion chamber [Mardani and Rosli, 2001: Suga et al., 1997).

Pre-mixing of natural gas and air in a NGV engine vehicle is inadequate because it adversely influences the engine combustion and emission characteristics of the exhaust gas. Therefore it creates some problem in determining how good the gas mixer would be in order to mix the air with the injected natural gas. It is crucial to determine the best air-fuel mixture ratio to ignite in the internal-combustion engine so that the engine would give the optimum performance and reduce fuel consumption as well as having cleaner exhaust emissions.

The aim of this study was to determine the performance of the NGV motorcycle mixer through experimental approach. Two types of NGV motorcycle mixer involved in this study are Mixer A and Mixer B. These two mixers were installed on a four-stroke motorcycle engine of a Modenas Kriss 110 cc model. The fuel used in this natural gas motorcycle system was Compressed Natural Gas (CNG). The experimental parameters involved in this study were flow rate of the natural gas, engine speed and torque, and the concentration of CO2, CO, HC, and NOx in the exhaust gas. This study basically covered the following areas: i. The basic concept of using the NGV mixer in the air-fuel

mixing system for the NGV system; ii. The engine power output analysis of the natural gas

A Study on Air-Fuel Mixing System for NGV Motorcycle System 33

motorcycle engine when two types of mixers were installed at different time;

iii. Exhaust emissions analysis due to the engine combustion when these mixers were used in the natural gas motorcycle engine system;

EXPERIMENTAL

All the equipment and apparatus used in this experiment especially the motorcycle engine must be at the standards state according to the manufacturer of the motorcycle. The motorcycle engine have been installed with a CNG conversion kit to adjust the engine operating condition in order to make the engine able to be run with both gasoline and natural gas. All equipment used in this experimental study is listed in Table 2.1.


Table 2.1 List of Equipment Used In the Experiment

Engine Type 4-Stroke Single Cylinder

Displacement 111 cc Valve System SOHC 2 valves Ignition System DC-CDI Starting System Electric and Kick Cooling System Air Cooled Engine Oil Capacity

1.1 litres

Transmission 4 Speed Return (Down)

Clutch Type Centrifugal and Wet Multi Disc

Maximum Power Output

9.0 PS (6.6 kW) @ 8500 rpm

Maximum Torque

0.95 Kgfm (9.3 Nm) @ 4000 rpm

Four-Stroke Motorcycle Engine

The four-stroke motorcycle engine that is used in this experiment is an engine from Modenas Kriss 110cc. The actual engine has been modified to run in the bi-fuel engine by installing a CNG conversion kit. Table 2.2 below shows the test engine specification of the Modenas Kriss 110 cc (Test Engine Specification: Modenas Kriss 1 Operating Manual).


Table 2.2 Test Engine Specification

Engine Type 4-Stroke Single Cylinder

Displacement 111 cc Valve System SOHC 2 valves Ignition System DC-CDI Starting System Electric and Kick Cooling System Air Cooled Engine Oil Capacity

1.1 litres

Transmission 4 Speed Return (Down)

Clutch Type Centrifugal and Wet Multi Disc

Maximum Power Output

9.0 PS (6.6 kW) @ 8500 rpm

Maximum Torque

0.95 Kgfm (9.3 Nm) @ 4000 rpm

Gas Mixers

Two types of natural gas motorcycle mixers that are involved in this study are: i. Mixer A: natural gas mixer with 4 holes; ii. Mixer B: natural gas mixer that has been modified from motorcycle carburetor.


Mixer A Mixer B

Figure 2.1 Two types of natural gas mixers

Parameters Involved In Determining the Engine Performance

Stated below were the parameters which can bring effect to the engine performance: i. Natural gas flow rate; ii. Engine speed; iii. Engine’s torque; iv. Concentration of CO2, CO, HC, and NOx in the exhaust

emissions.


Experimental Methods

There were two experimental approach has been used in order to determine the performance of both types of natural gas mixers. The first approach was by obtaining the engine power output and the second approach was by analyzing the exhaust emission to the environment when these natural gas mixers were installed to the test motorcycle engine.

In determining the engine power output, the engine was coupled with an eddy-current dynamometer (Figure 2.2). The torque of the eddy-current dynamometer and the engine speed were displayed at the dynamometer controller when the engine was tested. The procedures are as follow; i. Before the engine was started, Mixer A was installed and

tighten at the right position (replacing the motorcycle engine’s carburetor) of the motorcycle engine;

ii. After the motorcycle engine was started, it was let to run until it reached the steady state condition normally 20 to 30 minutes;

iii. Gas flow to the engine was fixed by controlling the gas control valve in order to determine the engine’s torque. The first gas flow rate used was 12.0 L/min;

iv. First gear transmission was used to run the eddy-current dynamometer;

v. The engine was run until the dynamometer controller gave a stable reading of torque and engine speed;

vi. The torque and the engine speed were recorded into a table; vii. The above 4-6 steps were repeated with different

transmission gear; viii. By controlling the gas control valve, gas flow rate were

changed to 12.5 L/min, 13.0 L/min, 13.5 L/min and 14.0 L/min;


ix. The above 4-6 steps were repeated for each of the gas flow rate;

x. After testing the first mixer, the engine was let to be cooled to ambient temperature;

xi. The entire steps were repeated using Mixer B.

Figure 2.2 Eddy-Current Dynamometer

There were two methods used in undertaking this exhaust emission analysis which was including running the motorcycle engine without load or idling test method and non-idling test method.


Idling Test Method

In idling test method the engine was run without any gear transmission and the gas flow to the engine were fixed to five different gas flow rate starting with 12.0 L/min until 14.0 L/min. At each of the gas flow to the engine the gas emission analyzer was put into a hole made at the exhaust system where the exhaust emission can be measured. The exhaust emission measured are the concentration of CO2, CO, HC, and NOx. i. Before the engine was started, Mixer A was installed and


ii. After the motorcycle engine was started, it was allowed to run until it reached the steady state condition normally 20 to 30 minutes;

iii. The engine’s torque was increased constantly by controlling the gas control valve for 15 seconds and then the gas control valve was controlled to release 12.0 L/min of natural gas to the engine;

iv. When the engine reached the idling state, the sensor of the gas emission analyzer was put into the exhaust flow;

v. The gas emission analyzer was set to record the exhaust emission after 40 seconds it is put into the exhaust flow. This is to ensure the reading is stable before it is recorded;

vi. The reading of exhaust emission which was recorded including CO2, CO, HC, and NOx;

vii. The above 3-6 steps were repeated for another different gas flow rate which was 12.5 L/min, 13.0 L/min, 13.5 L/min and 14.0 L/min;

viii. After testing the first mixer, the engine was let to be cooled to ambient temperature;

ix. The entire steps were repeated using Mixer B.


Non-Idling Test Method

For the non-idling test method the engine was run at a different gas flow rate starting with 12.0 L/min until 14.0 L/min. Each of the gas flow rate was used along with four different gear transmissions. At this condition, the concentration of CO2, CO, HC, and NOx in the exhaust emission was measured using the gas emission analyzer. i. Before the engine was started, Mixer A was installed and


ii. After the motorcycle engine was started, it was allowed to run until it reached the steady state condition normally 20 to 30 minutes;

iii. Gas flow to the engine was fixed by controlling the gas control valve. The first gas flow rate used was 12.0 L/min;

iv. First gear transmission was used in order to give the engine speed;

v. At this speed, the sensor of the gas emission analyzer was put into the exhaust flow;

vi. The gas emission analyzer was set to record the exhaust emission after 40 seconds it is put into the exhaust flow. This is to ensure the reading is stable before it is recorded;

vii. The reading of exhaust emission which was recorded including CO2, CO, HC, and NOx;

viii. The above steps were repeated for another reading at different transmission gear;

ix. The above 3-8 steps were repeated for different gas flow rate which was 12.5 L/min, 13.0 L/min, 13.5 L/min and 14.0 L/min;

x. After testing the first mixer, the engine was let to be cooled to ambient temperature;

xi. The entire steps were repeated using Mixer B.



Engine Power Output Analysis

Engine power indicates the actual usable power delivered at the engine crankshaft (Mardani et al., 2001). It depends on how much fuel is provided into the engine and it increases with the engine speed. A motorcycle engine is a mechanism designed to transform the chemical energy of burning fuel into mechanical energy. Thus motorcycle engine is an internal combustion engine. This means that the fuel that enters the engine’s combustion chamber must be combined first with air before it is burned inside the engine (Bo and Furuyama, 1996). The same trend of power against gas flow rate is illustrated in Figures 2.3, 2.4, 2.5 and 2.6.

Power (kW) vs Gas Flow Rate (kg/s) at First Gear

0.0000

0.0200

0.0400

0.0600

0.0800

0.1000

0.1200

0.1400

0.1600

0.1800

135.00 140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

Pow

er (k

W)

Power (kW) Mixer A Power (kW) Mixer B [Power (kW) Mixer A] [Power (kW) Mixer B]

Figure 2.3 Power against Gas Flow Rate at 1st Gears


Power (kW) vs Gas Flow Rate (kg/s) at Second Gear

0.0000

0.0500

0.1000

0.1500

0.2000

0.2500

0.3000

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

Pow

er (k

W)


Figure 2.4 Power against Gas Flow Rate at 2nd Gears

Power (kW) vs Gas Flow Rate (kg/s) at Third Gear

0.0000

0.0500

0.1000

0.1500

0.2000

0.2500

0.3000

0.3500

0.4000

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

Pow

er (k

W)


Figure 2.5 Power against Gas Flow Rate at 3rd Gears


Power (kW) vs Gas Flow Rate (kg/s) at Fourth Gear

0.0000

0.0500

0.1000

0.1500

0.2000

0.2500

0.3000

0.3500

0.4000

0.4500

0.5000

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

Pow

er (k

W)


Figure 2.6 Power against Gas Flow Rate at 4th Gears

It shows that Mixer B gives more engine power output as compared to Mixer A. The high power output achieved by the engine when Mixer B was used due to the effective mixing of fuel and air before the mixture entered the engine’s combustion chamber. Thus, it provides the engine with a better mixing of air-fuel that ignited and gave the engine more power output as compared to Mixer A. The overall result depicted was that Mixer B gave 48 to 59 % higher engine power output than Mixer A.

Exhaust Emission Analysis

This section discusses the effect of perfect mixing of fuel and air to the exhaust emission emitted by the motorcycle engine. As has been described, there were two types of exhaust emission test conducted in this experiment. The tests include the idling test method and the non-idling method. The concentration of carbon


dioxide (CO2), carbon monoxide (CO), unburned hydrocarbons (UHC) and nitrogen oxide (NOx) in the exhaust emission was analyzed.

Idling Test Method

In this method, the exhaust emission was measured when the motorcycle engine was run without load or any gear transmission with five different gas flow rate. The results of the test on the two mixers are illustrated in Figures 2.7, 2.8, 2.9 and 2.10.

Concentration of CO2 (%) vs Gas Flow Rate (kg/s) without Load

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

CO

2 (%

)

CO2 (%) Mixer A CO2 (%) Mixer B [CO2 (%) Mixer A] [CO2 (%) Mixer B]

Figure 2.7 CO2 against Gas Flow Rate


Concentration of CO (%) vs Gas Flow Rate (kg/s) without Load

1.55

1.60

1.65

1.70

1.75

1.80

1.85

1.90

1.95

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

CO

(%)

CO (%) Mixer A CO (%) Mixer B [CO (%) Mixer A] [CO (%) Mixer B]

Figure 2.8 CO against Gas Flow Rate)

Concentration of HC (ppm) vs Gas Flow Rate (kg/s) without Load

200

400

600

800

1000

1200

1400

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

HC

(ppm

)

HC (ppm) Mixer A HC (ppm) Mixer B [HC (ppm) Mixer A] [HC (ppm) Mixer B]

Figure 2.9 UHC against Gas Flow Rate


Concentration of NOx (ppm) vs Gas Flow Rate (kg/s) without Load

144

145

146

147

148

149

150

151

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

NO

x (pp

m)

NOx (ppm) Mixer A NOx (ppm) Mixer B [NOx (ppm) Mixer A] [NOx (ppm) Mixer B]

Figure 2.10 NOx against Gas Flow Rate

Figure 2.7 illustrates the concentration of CO2 against the gas flow rate. Figures 2.8 and 2.9 show the concentration of CO and hydrocarbons (HC emissions) against the gas flow rate respectively. Figure 2.10 is the concentration of NOx released by the engine with Mixer B which is much higher than the concentration released by the engine with Mixer A. The results show that Mixer B is more effective than Mixer A in producing turbulent flow for the fuel and air in order to well mix the fuel-air before entering the engine’s combustion chamber. As Mixer B has better mixing performance, it produces high concentration of CO2and NOx (due to high temperature) but low the CO and HC concentration in the exhaust emissions. For the range of the gas flow rate that was used in this experiment it is noticed that when the engine was operated without load, Mixer B was able to give better mixing quality than Mixer A. Thus, Mixer B is more suitable to be installed to the motorcycle engine than Mixer A at idling state.


Non-idling Test Method

As for non-idling test method, the concentration of the exhaust gas obtained is also plotted into graphs in order to simplify the comparison. The plotted graphs depicted the concentration of the exhaust gas emission against gas flow rate (Figures 2.11, 2.12, 2.13, and 2.14).

Concentration of CO2 (%) vs Gas Flow Rate (kg/s) at First Gear

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

CO

2 (%

)


Figure 2.11 CO2 against Gas Flow Rate at 1st Gear


Concentration of CO2 (%) vs Gas Flow Rate (kg/s) at Second Gear

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

CO

2 (%

)


Figure 2.12 CO2 against Gas Flow Rate at 2nd Gear

Concentration of CO2 (%) vs Gas Flow Rate (kg/s) at Third Gear

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

CO

2 (%

)


Figure 2.13 CO2 against Gas Flow Rate at 3rd Gear


Concentration of CO2 (%) vs Gas Flow Rate (kg/s) at Fourth Gear

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

CO

2 (%

)


Figure 2.14 CO2 against Gas Flow Rate at 4th Gear

The results show that at low gas flow rate, the released of CO2 for Mixer B is higher than Mixer A. This is because Mixer B would give better mixing of fuel and air before the mixture entered the engine’s combustion chamber than Mixer A. Thus contribute to the released of more CO2 from a complete combustion.

The same pattern is illustrated in Figures 2.15, 2.16, 2.17, and 2.18 for every gear transmission. The concentration of CO released by the engine with Mixer A was higher than Mixer B at low gas flow rate. This is due to poor air-fuel mixture or less air entered the engine’s combustion. However, at higher gas flow rate Mixer A was able to mix the air-fuel mixture more effectively thus lower the amount of CO produced. For Mixer B, at low gas flow rate, the mixing performance was better than Mixer A. However, as the gas flow increased, the released of CO also increased and it is slightly higher than the Mixer A although the concentration of CO was starting to decrease for both mixers.


Concentration of CO (%) vs Gas Flow Rate (kg/s) at First Gear

1.60

1.65

1.70

1.75

1.80

1.85

1.90

1.95

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

CO

(%)


Figure 2.15 CO against Gas Flow Rate at 1st Gear

Concentration of CO (%) vs Gas Flow Rate (kg/s) Fourth Gear

1.70

1.75

1.80

1.85

1.90

1.95

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

CO

(%)


Figure 2.16 CO against Gas Flow Rate at 2nd Gear


Concentration of CO (%) vs Gas Flow Rate (kg/s) at Third Gear

1.65

1.70

1.75

1.80

1.85

1.90

1.95

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

CO

(%)


Figure 2.17 CO against Gas Flow Rate at 3rd Gears

Concentration of CO (%) vs Gas Flow Rate (kg/s) at Second Gear

1.60

1.65

1.70

1.75

1.80

1.85

1.90

1.95

2.00

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

CO

(%)


Figure 2.18 CO against Gas Flow Rate at 4th Gear


UHC Emissions

For UHC emissions, the same trend is described for the concentration of hydrocarbons (HC emissions) against the flow rate for every gear transmission. Based on Figures 2.19, 2.20, 2.21, and 2.22, at low gas flow rate, the concentration of HC released from the engine with Mixer B was lower than Mixer A and it kept decreasing when the gas flow rate was increased. Again, this also shows that Mixer B is more effective in producing turbulent flow for the fuel and air to mix well before entering the engine’s combustion chamber. In contrast, the concentration of HC from Mixer A increased when the gas flow rate increased. This means that the air-fuel mixture in the Mixer B is more efficient than Mixer A, and the flame front is able to reach the level of complete combustion.

Concentration of HC (ppm) vs Gas Flow Rate (kg/s) at First Gear

200

400

600

800

1000

1200

1400

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

HC

(ppm

)


Figure 2.19 UHC against Gas Flow Rate at 1st Gears


Concentration of HC (ppm) vs Gas Flow Rate (kg/s) at Second Gear

200

400

600

800

1000

1200

1400

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

HC

(ppm

)


Figure 2.20 UHC against Gas Flow Rate at 2nd Gear

Concentration of HC (ppm) vs Gas Flow Rate (kg/s) at Third Gear

200

400

600

800

1000

1200

1400

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

HC

(ppm

)


Figure 2.21 UHC against Gas Flow Rate at 3rd Gear


Concentration of HC (ppm) vs Gas Flow Rate (kg/s) at Fourth Gear

200

400

600

800

1000

1200

1400

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00

Gas Flowrate (kg/s)

HC

(ppm

)


Figure 2.22 UHC against Gas Flow Rate at 4th Gear

NOx Emissions

As shown in Figures 2.23, 2.24, 2.25, and 2.26, the concentration of NOx released by the engine with Mixer B was much higher than the concentration released by the engine with Mixer A in every gear transmission. This explained that when the gas flow rate increases, Mixer B produces a higher mixing performance of fuel and air than to Mixer A, thus producing high temperature when the mixture is burned inside the engine’s combustion chamber. As a result, high temperature generated will lead to high NOx formation. For Mixer A, the situation is quite different because the mixing of fuel and air was not as efficient as Mixer B thus it gives lower temperature than Mixer B, hence the possibilities of less NOxformation.


Concentration of NOx (ppm) vs Gas Flow Rate (kg/s) at First Gear

143

144

145

146

147

148

149

150

151

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

NO

x (pp

m)


Figure 2.23 NOx against Gas Flow Rate at 1st Gear

Concentration of NOx (ppm) vs Gas Flow Rate (kg/s) at Second Gear

145.5

146

146.5

147

147.5

148

148.5

149

149.5

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

NO

x (pp

m)


Figure 2.24 NOx against Gas Flow Rate at 2nd Gear


Concentration of NOx (ppm) vs Gas Flow Rate (kg/s) at Third Gear

144

145

146

147

148

149

150

151

152

153

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

NO

x (p

pm)


Figure 2.25 NOx against Gas Flow Rate at 3rd Gear

Concentration of NOx (ppm) vs Gas Flow Rate (kg/s) at Fourth Gear

144

145

146

147

148

149

150

151

152

140.00 145.00 150.00 155.00 160.00 165.00 170.00 175.00 180.00 185.00

Gas Flowrate (kg/s)

NO

x (p

pm)


Figure 2.26 NOx against Gas Flow Rate at 4th Gear

From the exhaust emissions analysis, it can be concluded that Mixer B produced 6 to 30% higher concentration of CO2, 1 to


5% lower concentration of CO, 5 to 35% lower concentration of HC and 0.6 to 2% higher concentration of NOx than Mixer A due to the better mixing performance of air and fuel.

CONCLUSION

The effectiveness of a natural gas mixer to provide a good quality mixing of air and fuel can be determined by determining the engine performance and the exhaust emissions that emits by the NGV motorcycle system. Through the experimental approach, two different types of natural gas mixer is compared and one of the mixer was able to give more engine power output and released lower concentration of CO and HC but higher concentration of CO2 and NOx in the exhaust gas. The results showed that Mixer B is more efficient in mixing both air and fuel before the mixture is forced to enter the engine’s combustion chamber than Mixer A. Therefore, this result shows that the air-fuel mixing system plays an important role on the performance of a NGV engine and at the same time lower exhaust emissions to the atmosphere, especially CO and HC emission.

REFERENCES

Akira, S., Koji M., Tetsuro K. and Hiroshi, N. (1999). Reduction of Exhaust Emissions and Fuel Consumption on CNG Engine for Light Duty Truck. Mitsubishi Motors Corporation. SAE Transaction Paper. 9540453.


Bo, Y.X. and Mikio Furuyama (1996). Visualization of Natural Gas-Air Mixing Flow In The Mixer of A CNG Vehicle. 1-4.

James, A.G. and Rinchad, W.B., Small Gas Engine. 2nd Edition. Principles of Combustion. 5; 1979.

Mardani, A.S. and Rosli, A.B. (2001). The Comparison Study On 1.5 L Engine Performance And Emmission Using Gasoline and Natural Gas Fuel. Malaysian Science and Technology Congress 2001.

Mardani, A.S., Rosli, A.B. and Sin, K.L. (2001). Effects of Fuel Density on The Performance of A CNG Fuelled Engine. Universiti Teknologi Malaysia.

Rosli, A.B., Mardani, A.S. and Sin, K.L. (2001). Effects of New Mixer Dimension and Quality of Gaseous Fuel Inlet In A Compressed Natural Gas (CNG) Engine (Part 2). Universiti Teknologi Malaysia.

Rosli, A.B. and Mardani, A.S. (2001). The Realization of Optimum CNG Engine: It’s Implication on Engine Design. Advances in Malaysian Energy Research. UKM; October 29.

Suga, T., Knight, B. and Arai, S. (1997). Near-Zero Emissions Natural Gas Vehicle, Honda Civic GX. SAE Transaction Paper 972643.

Test Engine Specification. Modenas Kriss 1 Operating Manual. Zulkifli, A.M., Zulkefli, Y., Rahmat, M., and Martin, P.K.P.

(1999). Cleaner Emission From Natural Gas Powered Motorcycle. Universiti Teknologi Malaysia.

3FINITE ELEMENT STRUCTURAL

ANALYSIS ON A SINGLE STEP NGVM PRESSURE REGULATOR BODY

Rahmat Mohsin Khairul Sozana Nor Kamarudin

INTRODUCTION

Vehicular emissions are the main contributor to urban air pollution. Recent years have seen social demands for reduced automobile exhaust emissions and alternative fuels available from energy resources. The impact of vehicular emissions can be considered in terms of health and environmental risks; that the pollutants cause when introduced into the air (Morawska, 2003). Emissions of green house gases (GHGs) are generally seen as a large problem since a temperature rise caused by the increasing concentrations of GHGs in the atmosphere is likely to influence global climate (Hekkerta, 2005).

The use of natural gas as a transportation fuel can offer emissions and environmental benefits, energy diversity and energy security. Natural gas (NG) is a mixture of difference gases which consists primarily of methane with minor amounts of ethane, propane, butane and pentane. In many Natural gas deposits, methane makes up to 80 to 90 percent of gas (Gas Malaysia Sdn. Bhd, 2003). It is a hydrocarbon, which is colorless, odorless and much lighter than air.

Natural gas vehicles (NGVs) operate similarly to traditional vehicles, but it use natural gas fuel. According to Ontario standard


natural gas is compressed to 3600 pounds per square inch (psi), and enters the vehicle through the natural gas fill valve (receptacle). The dynamic range of compressed natural gas vehicles (NGVs) systems is 80:1 (248-3.4bar) for supply pressure and 160:1 (17.3 – 0.11g/sec) for fuel flow. The gas goes through the high-pressure line and enters the engine compartment. Gas enters the regulator, which reduces pressure from up to 3600psi to near atmospheric pressure. The vast majority of these NGVs are after-market conversion using three stages of pressure regulation and a venturi air / fuel mixer. With very accurate zero pressure differential regulation, one can obtain low emissions; good power, excellent drivability and install-and-forget durability. A zero pressure regulators should operate both at sonic velocity and in a turbulent flow regime for as much this dynamic range as possible.

In this project a pressure regulator prototype for Natural Gas Vehicle Motorcycle is set out to be verified using Finite Element Analysis (FEA). Finite element analysis is an indispensable analysis tool for engineers and designer in a broad range of industries, especially in automotive and aerospace industries. Project implementation of natural gas as a driving fuel for motorcycle (NGVM) has emerged in UTM, since the establishment of NGVM prototype by Gas Technology Centre in 2000. Pressure regulator is one of the specific components in NGVM. This component plays an important role, which is to decrease and control the pressure level to a suitable pressure range. This research is focused upon the development of pressure regulator system for Natural Gas Vehicle-Motorcycle (NGVM). The main objective is to designed new prototype of NGVM and gain the optimize thickness of this designed prototype of NGVM pressure regulator via the assistance of finite element analysis (FEA).

Finite Element Structural Analysis on a Single Step NGVM 61

NATURAL GAS VEHICLES

Development of Natural Gas Vehicles

The first natural gas engine was built in 1860, before the Natural gas vehicles have been used with much success in the United States since the 1960s and in Europe for nearly 50 years. Natural gas as a vehicle fuel has a long and established record in Europe, Canada, New Zealand, Australia, and in the U.S.A. In fact, there are currently more than 30,000 natural gas vehicles on U.S. roads and over 700,000 worldwide. The use of natural gas as vehicular fuel is growing more popular by the hour, as it is demonstrated by the fact that more than 60 countries have chosen this alternative fuel. In Malaysia, the number of registered vehicles is 12 million with 51% of them are using gasoline (Ministry of transport, Malaysia, 2002) and by July 2005, there are only 15,600 conventional NGV, which are mainly taxi.

Natural gas is chosen as a means of drastically reducing social health costs due to the reduction in polluting emissions, mainly particulates, benzene and ground level ozone. At the other hand NGV is chosen as an economic instrument, by taking advantage of its low cost in order to motorise expansion through cheap freights and increased savings for large sectors of society; at the same time exportable surplus of oil and fuels generate solid incomes to the nation (Goldin and Manconi, 2004)

Advantages of Natural Gas Vehicles (NGV’s)

NGVs operate so cleanly because the fuel is inherently clean. Natural gas is generally composed of at least 90 percent methane and may contain other hydrocarbons in small amounts including


ethane, propane and butane. Table 2.1 showed the composition of Natural Gas characteristic by Gas Malaysia (2003).

Table 3.1 Composition of Natural Gas Characteristic

COMPONENT MOL % Methane,C1 92.73 Butane, C2 4.07Propane, C3 0.77Isobutane,iC4 0.08n-butane, nC4 0.06Other Hydrocarbon 0.01 Nitrogen, N2 0.45Carbon Monoxide, CO2 1.83

Compressibility 0.9977

Specific Gravity 0.61 Density 0.7478 kg/m3

Molecular Weight 17.4518 Gross Caloric Value 9530 Kcal/Sm3

Burning Velocity (m/s) 0.3 Upper Flammability Limit 15.4 Lower Flammability Limit 4.5 Auto Ignition Temperature (oC) 640 Theoretical Air Requirement (m3/m3) 9.74

(Source: http://www.gasmalaysia.com)


Because of methane is relatively pure component, the emissions of hydrocarbon, carbon monoxide and in some cases, nitrogen dioxide can be significantly less from the NGV than from a gasoline or diesel vehicle. Per unit of energy, natural gas contains less carbon than any other fossil fuel, and thus lower carbon monoxide (CO) emissions per vehicles mile travelled. In addition, natural gas has an octane rating of 130, which means it has an efficiency advantage over gasoline (Rashidi, 2001)

Recently Hamid and Ahmad (2002) presented a comparison of the NGV and gasoline base engine performance where they found the volumetric efficiency of the NGV engine is reduced by about 15% and overall performance lowered by circa 9% at maximum torque and maximum power conditions. The substantial advantage that CNG has in anti knock quality is related to the higher auto ignition temperature and higher octane number compared to that of gasoline as shown in Table 3.2. Due to such antiknock properties, dedicated SI CNG engines could potentially be designed with compression ratio (CR) as high as 13:1 (Thomas and Staunton, 1999).


Table 3.2 Combustion Related Properties of Gasoline & CNG

Properties Gasoline CNG Motor octane numberMolar mass (kg/mol) Carbon weight fraction (mass %) (A/F)xStoichiometric mixture density (kg/m3)Lower heating value (MJ/kg)Lower heating value of stoic. mixture (MJ/kg)Flammability limits (vol% in air)Spontaneous ignition temperature (˚C)

80 -90 11087

14.61.3843.62.83

1.3-7.1480-550

12016.04

7516.791.24

47.3772.725-15645

Note: (A/F)x = Stoichiometric air fuel ratio

Natural gas has high ignition temperature of about 1,200 ˚F, as compared to gasoline with about 600 ˚F. It also has a narrow range of flammability, the concentration in air should be between 5 – 15 %. Below and above the limit, natural gas will not burn. The high ignition temperature and limited flammability range make accidental ignition or combustion of natural gas unlikely to occur. An American Gas Association study reported the injuries or fatalities after more than a half billion miles driven with natural gas vehicles.

In addition, natural gas vehicle (NGV) is also much safer than gasoline-powered vehicles. The fuel storage cylinders used in NGVs are much stronger than gasoline fuel tanks. The design of NGV cylinders subjected to a number of federally required “severe


abuse” tests such as heat and pressures extremes, gunfire, collisions and fires. NGV fuel systems are “sealed” which prevents any spills or evaporative losses. Even if a leak were to occur in the NGV fuel system, the natural gas would dissipate into the atmosphere instead of forming a spreading pool or vapor cloud on the ground, as other fuels do because it is lighter than air.

Natural Gas Vehicles (NGV) System

The engine compartment contains the regulator where the pressure of the natural gas is reduced from 3000 psi to the engine supply pressure. From the regulator the natural gas goes to the air/fuel mixer located on the intake manifold, meters the flow of gas according to the requirement of the engine in order to ensure optimum carburetion in terms of driving, consumption and emissions. This is represented by vacuum generated in the mixing devices. In fuel injected vehicles the natural gas enters the injectors at relatively low pressure (up to about 6 bars).Finally, the natural gas flows into the engine’s combustion chamber and is ignited to power the vehicle. Special solenoid-operated valves prevent the gas from entering the engine when it is shut off. There are six main component parts fitted to the vehicle, which are:

i. The filler connection which incorporates a non return valve.

ii. The NGV storage cylinder together with a cylinder valve which incorporates a fusible plug.

iii. A first stage regulator to reduce the pressure from 20MPa to approximately 0.7MPa.

iv. A vacuum operated NGV fuel lock-off valve, which prevents the flow og NGV fuel when the engine stops.


v. A second stage regulator which further reduces the pressure

vi. An air/gas mixer that measures air flow and meters the flow of gas into the engine.

NGV Technology

There are some recent important advances in NGV technology that will keep the industry on track with the most advanced technologies being produced by the major automotive manufacturers. NGVs now are compatible with computerized fuel injected engines. They are superior to carbureted vehicles because natural gas is injected directly into the combustion chamber in its gaseous state without having to go through a special gas/air mixer. This makes the changeover instantaneous from gasoline to natural gas and back again.

The newest systems are 'closed loop'; they are part of the systems that include oxygen sensors in the vehicle tailpipe, and provide feedback to the engine control systems to alter the fuel/air ratio depending upon the requirements of a vehicle's performance at any given time (Rashidi, 2005).

Natural Gas Vehicles (NGV) Pressure Regulator

The function of pressure regulator in Natural Gas Vehicles system is to reduce natural gas pressure from fuel cylinder and to manage gas pressure to engine fuel system. The pressure regulator must respond quickly to the changing of gas flow and have very predictable output pressure throughout range of flow, temperature and tank pressure. Besides that, the pressure regulator also


manages varying gas compositions through the discussion of Joule-Thomson effect, requirement of heat to prevent icing inside the regulator and resists damage from compressor oils. In addition, it also enables options for pressure sensor, high-pressure (HP) and low-pressure (LP), gas fittings and relief valve connection.

Typical NGV Regulator System

The electronic control device to reduce the natural gas pressure will allow a regular flow of gas every time the engine requires it. It is equipped with three natural gas reduction stages that allow stability at both high and low pressures and a high pressure solenoid valve upstream from the first stage. The absorption of heat, taken from parts of the regulator heated with the liquid of the engine cooling circuit, prevents the natural gas freezing during the fall in pressure phase.

Figure 3.1 TN I SIC Regulator produced by Landirenzo (Source:http://www.landi.it/eng/prodotti/scheda_prod_met536116000.html)


The flow of gas necessary for engine idling has a positive pressure from the second stage and is activated by means of a gas pipe separated from the main flow. It includes an electronic starting device with a built-in safety system that trips and shuts off the gas solenoid valves if the engine is switched off or even stalls. Figure 3.1 shows the figure of TN I SIC Regulator produced by Landirenzo.

Finite Element Analysis (FEA)

Finite Element Analysis (FEA) was first developed in 1943 by R. Courant, who utilized the Ritz method of numerical analysis and minimization of variation calculus to obtain approximate solutions to vibration systems. Shortly thereafter, a paper published in 1956 by M. J. Turner, R. W. Clough, H. C. Martin, and L. J. Topp established a broader definition of numerical analysis. The paper centered on the "stiffness and deflection of complex structures".

Clough introduced the name “finite element method” first in 1960. Before that, at Boeing and elsewhere, the method was called “direct stiffness method”. The finite element method is most widely used for analysis of stress and displacement in bodies under static load. Nevertheless, problem of buckling, vibration, and dynamic response may also be solved, and the method has been extended to handle the nonlinear problems of large displacement, plasticity and creep.

EXPERIMENTAL

Structural analysis on pressure regulator of Natural Gas Vehicle-Motorcycle (NGVM) implemented through finite element analysis


(FEA). The finite element method uses a substitute structure, whose parts are in a sense, pieces of the actual structure. Our substitute structure is a finite element. Serious study of solid and contact element have to be conducted prior to continue with the modeling process.

Detailed drawing of the pressure regulator model is visualized in three dimensions, symmetric and asymmetric via AUTOCAD V2004 and SOLIDWORK V2005 computer programmed. The model then exported to MSC/NASTRAN to perform analysis of flow rate deformation and stress of regulator base. In addition, structure vibration also is analyzing in order to verify the structure body of NGVM pressure regulator. Comparison of the various results obtained has to be carried out to predict the characteristics of fluid flow in term of pressure reduction caused by various setting of valve and forces exist on the regulator wall.

The exact deformation and stress tensor provided by the selected material of construction is analyses when pressure 3000 psi applied to regulator base to authenticate the ability of the structure to resist the force excreted by the flowing gas. Satisfactory results using the latest NGVM pressure regulator model obtained indicates the final step of the process.

RESULTS AND DISCUSSIONS

Two Dimensional Modeling

Two dimensional model of regulator base is drawn to provide conceptual idea for three dimensional drawing. The model is drawn by using AutoCAD version 2004 computer programming as shown in Figure 4.1. The model will be transformed to solids views for the thickness and structure modification. This model is


drawn based on previous work in the Development of Regulator base project.

NGVM Pressure Regulator body divided into two parts, namely Part A and Part B. Each part is drawn separately in shape and dimensional. Part A of regulator body represents the regulator bowl base, and drawn with a projection that shows the regulator bowl base from the upper view. Overall observation of this view shown that the regulator bowl base has 93mm length and 70mm wide with one hollow cylinder for fuel inlet.

Part B of regulator body represents the regulator cap base and drawn from the bottom view. The base regulator cap has 70mm external diameter. Both of Part A and Part B of regulator base have four unit holes with 5.5mm diameter screw in each of its diagonal. It is designed to strongly coupled the bowl and cap of regulator body. The general design of this regulator has been made suitable to be installed on the NGVM system.

Figure 3.2 Two Dimensional Drawing of Part A and Part B


Solidified Modeling

Solid model for regulator base is developed by using SOLIDWORK program based on the conceptual ideal of two dimensional drawing. The solid model is drawn using the prototype system to analyse deformations and stresses. By structural modification on the critical region of this regulator the optimized thickness could be obtained. Initially, the thickness of the regulator base is reduced by 0.5mm until the optimum thickness of the prototype is obtained. The thickness reduction of this prototype is done to get the optimise thickness of original regulator base before structural modification is carried out. The evolution of Structural Design Modification via SOLIDWORK is shown in Figure 3.3

Structural modification on the critical region of regulator base is carried out in order to obtain the ideal structure design of NGVM pressure regulator. Evolution of the structure modification design is developed using SOLIDWORK program.2. Thus, results in a new design of NGVM pressure regulator. Modification has been done to the original base of the regulator by incorporating fins to the body of the regulator, to increase total surface area of the regulator. The fin plays an important role to absorb heat from the surrounding in order to decrease temperature drop during the fall in pressure phase. It is also preventing freezing of natural gas. It is shown that Part A of the regulator has been modified by adding rib and thickness at the top part of the structure. While, Part B regulator is modified by adding the thickness at the bottom part of the structure.


Figure 3.3 Evolution of Structural Design Modification via SOLIDWORK

The optimised thickness of the newly designed NGVM pressure regulator is obtained after the structural modification. All the solid model that has been drawn have to be saved by using IGES format before being exported to MSC. PATRAN program for the structural analysis. The analysis will subsequently based on separation of both parts.

B


MSC. PATRAN / NASTRAN ANALYSIS

The finite element analysis upon the deformations and stresses of the regulator base is carried out by using MSC. PATRAN/NASTRAN. The applied pressure at the inner surface of the pressure regulator base will force the elements to move from it origin. The analysis will show the displacement nodes. Modification is necessary at the critical region of the prototype design to ensure the safety and durability of the prototype when it is pressure at 3000psi (2.07E+07 Pascal) from CNG storage tank. Analysis will also take a serious consideration upon the maximum tensor stress caused by the applied pressure.

Result obtained through the FEA analysis could be used to optimise the design of single step NGVM pressure regulator thickness, in order to handle the applied pressure produces by changes in gas flow rate and deformation. For structural applications, the yield strength is usually more important property than the tensile strength. Once it is passed, the structure has deformed beyond acceptable limits. Yield strength is the pressure which a substance is capable of supporting without fracturing. In industry, yield stresses are usually not even approached because the applied stresses are kept well below the yield strength by a safety factor on the order of 1.5 to 2.0. Yield strengths for each material used are summarized in Table 3.4.

Table 3.4 Yield Strength for Each Material

Yield strength Material Psi ( x 108 Pa )

Stainless steel 304 76500 5.27 Brass 57000 3.93

Aluminum Cast Alloy 20000 1.65


Factor of safety (FS), also known as Safety Factor, is a multiplier applied to the calculated maximum load (force, torque, bending moment or a combination) to which a component or assembly will be subjected. Factor of safety takes into account the imperfections in materials, flaw in assembly, material degradation, and uncertainty in load estimates. An alternative way to use the safety factor is to derate the strength of the material to get a"design" strength.

FactorSafetyStrengthYieldStressDesign

On this analysis, the safety factor that will be considered is 1.5. Thus, the design stresses calculated for each material are shown in Table 3.5 and the optimise thickness of each material summarized as in Table 3.6.

Table 3.5 Material Design Stress

Material Design Stress ( x 108 Pa ) Stainless steel 304 3.51

Brass 2.62 Aluminum Cast Alloy 1.10


Table 3.6 Optimize Thickness of Each Material Selection

SS304 Brass Aluminum Cast Alloy Material

Construction Part A

Part B Part A Part B Part A

PartB

Stress Tensor( x 108

Pa ) 3.37 1.76 2.38 1.43 1.04 0.854 Deformation ( x 10-2

mm ) 1.89 0.504 3.40 0.876 3.38 1.08 Material Design Stress (x 108 Pa) 3.51 2.62 1.10 Optim.Thickness (mm)

3.0 3.5 6.0

Comparison on MSC/PATRAN Simulation Result

For the mass analysis, it base on the three type of the material selection, which is Stainless Steel (SS304), Brass (RED; 80% CU, 20% Zn) and Aluminum (pure). Even though, aluminum cast alloy give the largest thickness among the selected material construction but this material offer the minimum value of weight which is 63.32% less than SS304.

Decisions on selecting the most preferable material for the fabrication of this regulator also have to take serious consideration on the material market price, in order to gain the mot economic value. Based on the latest market price obtained from Walsin Precision Technology Sdn. Bhd. Company, Aluminum Cast Alloy gives the lowest market price per kg of material, followed by brass and stainless steel 304 as shown in Table 3.7.


Table 3.7 Gross Material Cost of Regulator New Prototype

Material Stainless Steel 304

Brass CastAluminum

Alloy Total Weight (gram)

1349.0302 1506.3597 494.80

Price (RM/kg) 33.918 19.80 17.00 Gross Cost (RM)

45.76 29.83 8.41

Therefore the gross cost to fabricate the new regulator base prototype when used aluminium as the construction material is RM8.41.While, by using Brass and Stainless Steel 304 as construction material the gross cost is RM29.83 and RM45.76. It is 71.81% and 83.9% higher than aluminium gross cost.

CONCLUSION

From this research project through the use of finite element analyses, the optimised design of NGVM pressure regulator thickness with the appropriate satisfying material construction is obtained. In the design of pressure regulator structure, it is primarily important to verify that the regulator is able to effectively manage NGV fuel system. This is to ensure the safety and durability of the pressure regulator when it operates at the applied pressure.

Cast Aluminum Alloy as construction material provide minimum amount of stress tensor compared to Brass and Stainless


Steel 304, which is 0.104 GPa occurs at node number 4681. Though the optimize thickness is 6.0mm, but total weight of base prototype using Cast Aluminum Alloy offer the minimum mass and gross cost.

REFERENCES

Thomas J.F. and Staunton R. H. (1999) SAE Tech. Pap. Seri. 1511(01):66.

Hamid HA, Ahmad AS. (2002). Development of monofuelled natural gas vehicles: a Malaysian experience. Malaysia: Special Project of Petronas Research. p.10.

Rashidi. (2001) Compressed Natural Gas Vehicles. Available from: http://succ.shirazu.ac.ir/~motor/ngv1.html

Landi Renzo, (2003). TN 1 SIC regulator. Available from: http://www.landi.it/eng/prodotti/schede_prod_met536116000.html

Gas Malaysia. (2003). Composition and Characteristic. Available from: http://www.gasmalaysia.com

Ontario (2005).About NGV. Available from: http://www.NGV Ontario.com

Diego Goldin and Santiago Manconi,(2004). NGV Technological Developments Buenos Aires NGV 2004

4EMISSION AND PERFORMANCE EVALUATION OF MOTORCYCLE

Zulkifli Abdul Majid Zulkefli Yaacob Rahmat Mohsin

INTRODUCTION

The importance of air pollution and environmental protection has drawn much attention in Malaysia since global environmental problems first emerged as a common worldwide concern at the United Nations Conference on Human Environment in 1972. In Malaysia, urban air pollution is reaching a critical level as witnessed during the recent haze crisis. The average emission of fine particulates is 77 g/m3. This figure is above the acceptable standard (50 g/m3) allowed by the Malaysian Department of Environment (Department of Environment, 1999). The urban transportation and industrial activities in Malaysia have aggravated the problem. The Federal Territory of Kuala Lumpur has the highest growth rate in the field of transportation, utilities and manufacturing activities in the country. The vehicle population has been increasing tremendously, thus aggravating the sulphur and lead content in the atmosphere. Motorcycle use has rapidly expanded over past several years, especially in the urban city such as the Federal Territory of Kuala Lumpur. In Malaysia, nearly five million units or over half of the motor vehicles in Malaysia are motorcycles. The increasing number of motor vehicles is from 4,692,183 in 1998 to 5,082,473


in 1999 with the highest concentration of vehicle population in the Federal Territory, Kuala Lumpur (Department of Transport Malaysia, 1999). The combustion of oil fuels leads to the emission of carbon monoxide, carbon dioxide, nitrogen oxides, sulfur dioxide and fine particles. Statistic in 1998 has shown that two million tones of carbon monoxide, 111 thousand tones of unburned hydrocarbons, 237 thousand tones of oxides of nitrogen, 38 thousand sulphur oxide and 17 thousand tones of particulate matters were emitted into the atmosphere (Department of Environment Malaysia ). These polluting substances lead to the production of smog that destroys sensitive tissues (in people, animals and plants), the formation of inhalable carcinogenic particles, reduced lung function, and ultimately responsible for many untimely deaths every year. Furthermore, air pollutant also trap excess heat that contributes to global warming and climate change, rising sea levels, changes in vegetation and increase in the frequency of severe weather events. To reduce air pollution caused by vehicle, Malaysia government support PETRONAS to introduce Natural Gas Vehicle (NGV) program by providing incentives (Hamzah and Abdul Shukor, 1996; Seisler, 1996) to the users. NGV, which produces much cleaner burning, will reduce the emission of hydrocarbon and carbon monoxide, which is normally produced by gasoline driven cars (Raine and Zoeliner, 1986; Arizipe et al., 1992; Suwasono, 1996; Akira et al., 1999). However, PETRONAS only introduced NGV system to the car and heavy vehicles (Hamzah and Abdul Shukor, 1996). Combustion of fossil fuels in an engine can be described as the chemical conversion of hydrocarbons and oxygen into water and carbon dioxide. During this conversion, thermal energy is released which, in turn is used to generate power. Next to carbon dioxide, internal combustion engines emit other greenhouse gases such as carbon monoxide (CO), oxides of nitrogen (NOX),

Analysis of Household Energy Demand 81

unburned hydrocarbon and non-methane hydrocarbon (NMHC). The formations of these greenhouse components in an internal combustion engine are the result of various chemical and physical changes during the combustion process (DeLuchi, 1987). Gasoline powered vehicles produce oxides of nitrogen (NOx) that when combined with volatile organic compounds (V0C’s) which are produced by trees naturally, will react with sunlight in the lower atmosphere to form ozone, a primary constituent of smog. CNG powered vehicles are clean - they emit 85% less NOx, 70% less reactive hydrocarbons, and 74% less carbon monoxide than similar gasoline powered vehicles. The use of CNG-fuel vehicles significantly reduces emissions of ozone precursors. A significant change has occurred in the energy policies of many nations throughout the world, including Malaysia. These new policies and the programs to implement them, herald the beginning of global transition away from oil as the dominant transportation fuel and toward the use of cleaner, more abundant and eventually sustainable energy resources (Brimblecombe Frances, 1995). Natural gas has proven to be cleaner, cheaper, safer and more domestically abundant than gasoline or other transportation fuels (Seisler, 1996). Propelled by favorable government policies and aided positive economic and environmental attributes, natural gas vehicles have an impressive growth in Malaysia. Bi-fuel motorcycle is a new technology introduced to alleviate emission problems and became as an option for the present system available with the usage of natural gas as alternative fuels. This motorcycle can run either with gasoline or natural gas. This type of motorcycle gives mono flexibility in term of fuel usage. As it has been establish that natural gas gives cleaner combustion and this paper will describes the emission and performance evaluation of the above said bi-fuel motorcycle.


MATERIALS AND METHODS

Experimental equipment

The motorcycle used for this study is MODENAS KRISS 110cc, 4-stroke single cylinder. The engine has been modified so that it can operate on either gasoline or natural gas. The specifications of the motorcycle are listed in Table 4.1.

Table 4.1 Specification of motorcycle

Type 4 st, 1 cyl, SOHC

Bore x stroke (mm) 53.0 x 50.6 Displacement (m3) 1.11 Compression ratio 9.3 Carburettor type KEIHIN PB18 X 1 Diameter of throttle valve (mm) 18 Diameter of venturi (mm) 18 Type of choke valve Butterfly Lubrication system Forced lubrication

Wet type Engine oil: Rating SF OR SG Viscosity (SAE Grade) 20W-40

Capacity (m3) 11 Cooling Method Air cooled Ignition system Magneto to CDI

6.5 BTDC /1200 ~ Ignition timing Angle( o/rpm) 27 BTDC / 4000 Spark plug type NGK C6HAS Gap (mm) 0.7


Exhaust emission tests standards

Emission test were carried out in accordance with the ISO 3929, ISO 6460 and ISO/TR 6970 test procedure. The exhaust emission for both gasoline and natural gas was analysed at idle speed and average speed of 40 - 90 km/hr. The motorcycle was tested on a chassis dynamometer at various constant speeds: 0, 40 to 90km/hr, respectively. Horiba MEXA 324J Infrared emission analyser was used to detect CO and HC emission while ENAREC 2000 emission analyser was used to detect NOx.

Test rig

The experimental rig comprise of an engine of MODENAS, KRISS 110cc motorcycle, a chassis dynamometer, emission analyzer and a data acquisition system. A data acquisition system, which is a data translation converter and an IBM computer, are used to record data such as engine speed, torque, power, exhaust temperature, and engine temperature. The test data is converted to standard operating conditions using ECE Code. The schematic diagram of the experimental equipment is shown in Figure 4.1.

Regularity C Air cleaner Wet element air filter

Number (qty) 1


Figure 4.1 Schematic diagram of the experimental equipment

Combustion Fuel

Natural gas and gasoline (PETRONAS, Primas PX2) has been used as a fuel to run the motorcycle. The composition of the natural gas and the specification of gasoline are shown in Table 4.2 and Table 4.3 respectively.


Table 4.2 Natural gas composition

Component Mol %

C6+ 0.07 C3 0.90 iC4 0.29 nC4 0.13 iC5 0.07 N2 0.68 C1 93.07 CO2 1.10 C2 3.70 Compressibility 0.9977 Density (at standards condition) 0.7404

kg/m3

Relative Density 0.6042 Molecular Weight 17.4663 Gross Calorie Value (at standard condition)

39.20 MJ/m3

Table 4.3 Gasoline specification (PETRONAS Primas PX2)

Description Value

Density @ 150C, kg/l 0.733 Research Octane Number (RON), g/l

97.0

Lead Content, kPa 0.008 Reid Vapour Pressure, %wt 62 Total Sulphur Trace Distillation 50% evaporated, 0C90% evaporated, 0C

105152

Colour Yellow



Carbon Monoxide (CO) emissions

It was found that at idle speed, CO emission from natural gas powered motorcycle was at an average of 0.02 % vol., which is equivalent to 99.7 % decrease for gasoline powered (3.998 %vol.). Whereas at constant speeds of 40 to 90 km/hr, the amount of CO from natural gas powered motorcycle was between 0.02-0.06 % due to good mixing of natural gas with air and consequently lead to complete combustion for natural gas compared with gasoline. Therefore, the percentage of CO emission is almost zero percent. Figures 4.4 and 4.5 illustrate the test results for idle speed exhaust emission and average CO emission respectively.

0.02 3.998

48.875

236.33

0

50

100

150

200

250

CO% HC ppm

Idle Speed Exhaust Emission

Natural Gas Petrol

Figure 4.4 Idle speed exhaust emission


Average C O Em ission

0

1

2

3

4

5

6

7

8

0 20 40 60 80 100

Average S peed (km /hr)

CO

vol

%

G asolineN G V

Figure 4.5 Average carbon monoxide emission

Unburned Hydrocarbon (UHC) emissions

The amount of unburned HC emitted by natural gas motorcycle was 79.3% lower than gasoline fuelled motorcycle, which is equivalent to 48.875 vol. ppm at idle speed (Figure 4.4). Similar result was obtained for constant speed of 40 km/hr to 90 km/hr where the natural gas powered motorcycle produces unburned HC at approximately 48 ppm as shown by Figure 4.6. The amount of UHC emitted from natural gas powered motorcycle is lower because of the readiness and ability of natural gas to mix very well with air in which that will ensure the completeness of the combustion.


A v e r a g e H C E m is s io n

0

5 0

1 0 0

1 5 0

2 0 0

2 5 0

3 0 0

0 2 0 4 0 6 0 8 0 1 0 0

A v e r a g e S p e e d (k m /h r )

HC

vol

. PPM

G a s o l in eN G V

Figure 4.6 Average hydrocarbon emission

Nitrogen oxides (NOx) emissions

Figure 4.7 illustrates the emission of NOX produce by bi-fuel motorcycle at different engine speed. The emission of NOX profile for gasoline showed the increment with the increasing of engine speed. Natural gas fuelled motorcycle is really effective to eliminate nitrogen oxides. According to the testing on bi-fuel motorcycle while using natural gas clearly shows that nitrogen oxides was totally eliminates when operates on natural gas compared to gasoline at different operation conditions. Figure 4.8 shows the relationship of engine oil temperature and emission of NOX. Emission of NOX is increase steadily with temperature. Engine oil temperature for gasoline-fuelled motorcycle is higher than the engine oil temperature for natural gas fuelled for the same motorcycle. It shows that the temperature in the combustion chamber is higher when using gasoline as a fuel. This is because gasoline has much higher calorific value compared to natural gas. The impact of this fact is more significant in relation of the formation of NOX. The higher operating temperature when


using gasoline encourage much easier for the formation of NOX as compared to when using natural gas.

0123456789

NO

x (p

pm)

0 0 175 246.15 316.48 421.97

Engine Speed (rpm)

NOx Emission at Different Engine Speed

Petrol

Natural Gas

Figure 4.7 NOx emission at different engine speed

0

10

20

30

40

50

60

NO

x (p

pm)

26.6 30 32 36 40.6 41.6

Engine Temperature

NOx Emission for Different Engine Oil Temperature

PETROLNATURAL GAS

Figure 4.8 NOx emission at different lubricating oil temperature


Combustion efficiency

Figure 4.9 illustrate the relationship of combustion for both, petrol and natural gas with the emission produced during the test on various engine speeds. At low engine speed, the combustion of petrol fuel gives unsteady and poor combustion efficiency. The combustion efficiency for this fuel increases as the engine speed increases. The combustion of natural gas fuel instead shows about constant trend. This is due to the characteristics of the fuel itself. Unlike petrol, which must be vaporized before ignition, natural gas is already in gaseous form when it enters the combustion chamber. As the intake valve opens, the gas enters the combustion chamber, where it is ignited easily to power the motorcycle. Figure 4.10 shows profile of combustion efficiency at different engine temperature for gasoline and natural gas. This figure shows the relationship of combustion efficiency with combustion temperature in the engine. Gasoline fuelled motorcycle shows unsteady combustion efficiency at different engine temperature. Natural gas as before, gives constant high combustion efficiency. These two figures (Fig. 4.9 and 4.10) confirm each other in term of combustion efficiency, where the combustion efficiency of natural gas fuel is better than gasoline fuelled motorcycle.


0102030405060708090

100

0 0 500 700 900 1200

Engine Speed (rpm x Gear Ratio)

Com

bust

ion

Effic

ienc

y (%

)

Petrol Natural Gas

Figure 4.9 Combustion efficiency at different engine speed

Figure 4.10 Combustion efficiency at different engine temperature

0

10

20

30

40

50

60

70

80

90

100

28 30 32 33 35 37 38 38 38

Temperature (0C)

Com

bust

ion

Effic

ienc

y (%

)

Petrol Natural Gas


Engine Oil Temperature

Figure 4.11 shows the engine oil temperature while running both gasoline and natural gas. Engine oil temperature for gasoline fuelled motorcycle is higher (85.5 oC) than the engine oil for natural gas fuelled for the same motorcycle. It shows that the temperature in the combustion chamber is higher when using gasoline as a fuel. This is simply because gasoline has much higher calorific value compared to natural gas. The impact of this fact is more significant. The higher operating temperature when using gasoline encouraged much easier for the formation of NOx as compared to when using natural gas as a fuel for the engine.


Figure 4.11 Lubricating oil temperature over running time

Figure 4.12 show the maximum engine power as a function of engine speed for both fuels. The results are obtained at wide-open–throttle with manufacturer setting ignition timing. During the operation using natural gas, the power was reduced by approximately 15% at high engine speed. This power loss is due to the displacement of air by natural gas and low burning velocity of natural gas compared to gasoline. In another hand, despite natural gas characteristics has high octane rating number of 130 compare to gasoline, which has only 92 to 98 but due to the bi-fuel system the compression ratio is still set for gasoline setting then the inherent performance benefits of natural gas which is high compression ratio are lost. As the engine speed increase, the engine

Engine Temperature Versus Running Time

010203040

5060708090

Running Time (Mi

GasolineNatural Gas


power is further decreased. This is due to the design of mixer, which is not suitable for high flow rate. Although natural gas is having high octane rating and the engine could operate at a compression ratio up to 16:1 without “knock” it is not possible because the motorcycle used for the test has only 9.3:1 compression ratio as its design to operate for gasoline. As the result, the natural gas powered motorcycle requires advance spark ignition timing compare to gasoline in order to utilise the high compression ratio and high octane rating characteristics of natural gas.

Engine Power At Wide Open Throttle

0

1

2

3

4

5

6

7

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

Engine Speed (1/min)

Engi

ne P

ower

(kW

)

PetrolCNG

Figure 4.12 Engine power against engine speed


CONCLUSION

The test result shows very favourable indication for the use of natural gas as a fuel for motorcycle. Although natural gas powered motorcycle has less power output but this newly developed technology is capable of eliminating nitrogen oxides and gives a significant decrease of CO and UHC emission with high combustion efficiency. This is because natural gas has already in gas phase and can mix with combustion air easily at the mixer. Due to that the combustion of natural gas give a complete burning of the fuel itself hence less exhaust gas emission emitted to the atmosphere. Since the air pollution is a major concern today and leads to annihilation of an environment and human health especially in densely populated area and congested city centres, the natural gas powered motorcycle is one of the measures to solve this problem.

REFERENCES

Akira S. Koji M., Tetsuro K., and Hirosi N. Reduction of Exhaust Emissions and Fuel Consumption on CNG Engine for light Duty Truck. Mitsubishi Motors Corporation. SAE Transaction paper 9540453, 1999.

Arizipe, L., R. Costanza and W. Lutz. “Population and Natural Resource use” In J. Dooge Al (Eds) “An agenda of Science for Environment and Development into the 21st Century”. London UK: Cambridge University Press; 1992.

Brimblecombe, Peter and Nicholas, Frances. “Urban Air Pollution and its Consequences” in Timothy O’Riordan (Ed),


Environmental Science for Environmental Management. London: Longman Group Limited, London; 1995.

DeLuchi M.A., “Transportation Fuels and the Greenhouse Effect”, Division of Environmental Studies, University of California, Davis; 1987.

Department of Environment Malaysia. A Guide to Air Pollution Index, Kuala Lumpur: DOE; 1999.

Department of Transport Malaysia, Annual Pollution Report: Kuala Lumpur: DOT 1999.

Djoke Suwasono. “Air Pollution from Movement Emission Source and its Problems”. Proceeding of the 5th Biennial IANGV International Conference on Natural Gas Vehicles, Kuala Lumpur, 1996.

Hamzah Abd. Hamid and Abdul Shukor Ahmad. “The Development of Monofuelled Natural gas Vehicles; A Malaysian Experience”. Proceeding of 5th Biennial IANGV International Conference on Natural Gas Vehicles; Kuala Lumpur, 1996.

Jeff M. Seisler, “NGV Politics: Opportunities for Governments to Promote Clean Fuels”. Proceeding of 5th Biennial IANGV International Conference on Natural Gas Vehicles, Kuala Lumpur, 1996.

Jones K., Raine R.R., Zoeliner S., “A Study of the Natural Gas Composition on Engine Performance, Emissions and Efficiency”. Conference on Gaseous Fuels for Transportation, Vancouver, 1986.

5WATER JETTING IMPACT ON

NATURAL GAS PIPING (ABRASION EFFECT AND METAL

THINNING)Zulkifli Abd Majid

Rahmat Mohsin Fadhli Omar

INTRODUCTION

Problem of erosion has long been an issue in the engineering world. Work on erosion by liquid impact has been studied by various body especially one related to the mechanical engineering field. Liquid impact upon mechanical equipments or metals can be seen in forms of cavitations or metal degradation due to erosion. Among pioneer problems that resulted to this study is the fact that buried natural gas piping are always under the threat of damage due to water jetting in cases where water utility piping which runs parallel to it got leaked. Degradation of piping material not only disrupts the continuous supply of natural gas it too could pose a threat to life and buildings in terms of lost of lives and destruction of buildings. Cases of natural gas piping failure has been proven to be catastrophic as demonstrated in cases of pipe rupture in South Riding, Virginia in 1998 and Carlsbad, New Mexico in 2000 (National Transportation Safety Board, 2000; National Transportation Safety Board, 1998). The problem of erosion has been discussed briefly in 1961 by S.M. DeCorso and R.E. Kothmann from Westinghouse


Research Laboratories, Pittsburgh in their paper entitled ‘Erosion by Liquid Impact’. In that paper it was concluded that erosion at high speed impact of greater than 700 feet per second can never be contributed by chemical action (DeCorso and Kothman, 1961). Therefore, when chemical action was eliminated the only possible causes that can be contributed to erosion of the blades are erosion due to mechanical action of the liquid impact during its operation.However a jet’s ability to incur enough damage in terms of thinning of the impacted surface highly depends on the force of the impact. Force of an impact will vary with the distance away from the jetting source as every jet will have its own effective jet length (DeCorso and Kothman, 1961). In short, a jet which is not in its effective distance will produced damages that is less than the damages caused by a well formed jet at the same velocity. This can be explained by the area of the leading edge of the jet. An irregular shape for a jet’s leading edge will not be able to produce an impact area which is as large as the one in its effective distance due to the area behind the leading edge has a reduced velocity compared to the leading edge. This happens as a result of momentum lost as the length travelled by the jet increased. Referring to general belief, one normally expects a higher rate of deformation and erosion of a material which has a rough surface due to cavities or irregularity of impact surface. Unfortunately, the opposite has been found to be the real answer. For example, a test by having the surface area of impact to be placed by a single diamond point indentation will give a result which without a doubt rendered surface irregularity merely a minor effect to the overall damage done by the liquid impact (DeCorso and Kothman, 1961). Another unique characteristic that can be seen after a material has been deformed due to liquid impact is that it creates a ring deformation. The main feature of the deformation is a ring crack separating a region of intense circumferential fracture from a central unfractured area of the surface (Brunton, 1961). Central


unfractured area of the surface happens due to the displacing properties of the liquid impact. During the leading edge impact on the surface pressure generated by the impact velocity results to the exertion of enough force onto the surface to exceed the materials shear strength. The force will be displacing particles on the surface continuously and uniformly thus creating the central unfractured area. Particles being displaced from the central unfractured area will be force to move upwards towards the opening created by the liquid impact. This movement at high velocity resulted to the outer ring damages through shearing or tearing of the material surface. For the particles from the central area to move upwards it has to be at a velocity which is higher than the impact velocity. In short, the outer ring was damaged due to the radial force which is grater compared to the impact force due to increase of outward flow velocity [4]. This could be explain by Equation 5.1 which stipulates an increase of impact velocity will result to a higher impact force thus contributing to a higher abrasion rate to a surface.

F = CVA (5.1)

where, F is force generated, is liquid density, C is velocity of sound in liquid, V is impact velocity and A is area of impact. Abrasion by liquid impact is termed as a situation when rigid abrasive particles transported by a liquid stream/jet remove material from an impacted surface. Erosion is termed as degradation of an impacted surface by means of solid particles in fluid stream. Ductile materials refer to material which undergoes erosion process in the form of plastic deformation. Volume removal is normally due to the cutting or the displacing action of the solid particle in the liquid jet. Firstly it is to be expected that impact perpendicular to the surface would produce indentations but little volume removal in ductile material. However it is hardly the


same for in cases where impact directions making a lower angle with the impacted surface due to the grooves which will be cut off the surface after the impact. Although it is true that little volume is removed when the impact is perpendicular to the impacted surface, one must not forget small cuts resulted from the impact can be aggravated when there are multiple impact as in the real situation of liquid jet. In brittle material an entirely different result will be seen after an abrasive liquid jet impact. This is due to the fact that in brittle material erosion does not happen by means of plastic deformation but it is more likely related to the material’s elastic behavior. Elastic stress of the impacting particles will contribute to the displacing action of the particle through propagation and intersection of cracks ahead of the impacting particles. It is found through a test that brittle material erodes most rapidly at angles near to 90 degree (Finnie, 1961). The explanation for this is that in brittle material angle of 90 degree for the impacting particle with the impacted surface allows a maximum component of velocity tangential to the surface with regard to angular particle. This maximum tangential velocity will be able to exert enough force to firstly propagate cracking of the impacted surface. This paper presents a study on abrasive water jet upon carbon steel piping, where through this study various aspects of abrasive water jet impact upon the sample can examined in terms of jet distance, abrasion effect and thinning rate through the usage of 5mm diameter orifice in water to sand system.


EXPERIMENTAL SET-UP

Sample Preparation

Firstly, the thickness of the coating and thickness of the pipe without the coating is measured by thickness measuring equipment. All the data for the coating thickness and the thickness of the pipe without the coating were recorded. Several sets of measurement data for every specimen were taken in order to get an accurate and minimum error in reading. Basically, the surface of the specimen pipe is divided into 4 horizontal sections with 5 horizontal lines, namely Line A, Line B, Line C, Line D and Line E. Next, the horizontal lines were separated into 3 areas with the data taking points being the center of the area. Coating thickness data on each area were acquired using digital coating thickness gauge. In order to determine metal thickness of each specimen a section of the specimen were scraped off its coating to reveal the bare metal surface of the specimen. Three data measurement were taken from the designated point as shown in Figure 5.1.

Figure 5.1 Data taking template


After all the procedure has been followed correctly, the specimen is tied to the pipe holder rig. Then, the position of the specimen is needs to be set in order to make sure the jetting of the orifice is parallel to the line and points that have been selected the surface of the pipe.

Experimental Setup

Basically a section of natural gas carbon steel pipe will be attached to the saddle on the rig. The saddle will be adjusted to a distance deemed suitable for the experiment. Next an orifice is attached to the orifice holder. Supply pressure of the water is to be set at 10barg while the distance away from the jetting source is varied at 10 cm till 70 cm. Next, sand that has been sieved to be at 600 -2000 m is filled into the handling tank before the rig is to be lowered into the tank. A number of small holes drilled on the rig structure will act as a creator of water jet which will be used to disperse sand in the tank to allow the rig to be lowered into the sand thus burying it. After the designated interval the rig will be transported out from the handling tank using a chain block. Thinning of the sample pipe material will then be measured on points dictated based on a data taking template. Figure 5.2 is a schematic diagram depicting the experimental set up.


Figure 5.2 Experimental set up

Determination of Abrasion Effect

This is done through visual inspection of the impacted surface after the end of testing which coincides which the identification of rupture point on the specimen surface. Besides visual inspection, physical examination too was administered through the act of touching the impacted surface with bare hands to determine the surface roughness.

Determination of Metal Thinning

Metal thinning effect was monitored through specimen thickness measurement after a period of jetting exposure. Thickness measurement gauges used in the experiment was of two types. First is the Digital Coating Thickness Gauge (Model TT260) used in measuring the thickness of the pipe coating. While the thickness of the bare pipe is measured using the Digital Ultrasonic Thickness Gauge (Model TT130).



Visual Analysis of the Impacted Surface

From physical examination it can be detected that orifice jetting produces an impact on the specimen pipe which resembles a round shape. One thing peculiar about the impacted surface is that the roughness of the impacted surface is not evenly scattered with one being smooth while the other being very rough. Figure 5.3 depicts a specimen surface after an exposure of 170 hours to abrasive jetting.

Figure 5.3 Specimen surface after impact

The one which has water as the main erosion media has a smoother surface when compared to the one which has sand as the main erosion media. Area 1 is the one that has water as the

Roughsurface Smooth

surface


majority erosion media (Figure 5.4). This gives an indication of the velocity of that area being higher compared to the velocity of Area 2 (Figure 5.4) which is normal in cases of orifice jetting where pressure at the jet’s centerline is higher compared to the edge of the jet structure (Greitzer et al., 2004). For example, velocity of the orifice outlet in Area 1 is a lot higher compared to the one in Area 2. Therefore its ability to carry sand particles is lesser due to the weight of the sand which reduces the sand particles’ ability to move with high velocity as in Area 1. Theoretically this could be verified with the equation below;

F = ma (5.2)

where, F is force, m is mass, a is acceleration

Figure 5.4 Orifice jetting profiling

From Equation 5.2, it is safe to agree upon the fact that an increase in mass will result to the reduction of the sand particles’

Impacted surface

Area 1

Area 2


ability to accelerate in order to achieve the sufficient force to be at the velocity needed to move in Area 1. Bearing this in mind it is not entirely misleading if it were to be concluded that the existence of two different category of area upon the impacted surface is due to the way of the sand particles’ dispersion and concentration in various area of the water jet.

Profiling of Jetting Dispersion

Dispersion of water jet structure away from its center point is basically due to the high resistance of the initial impacted media which in this case is water impacting on sand particles. This failure resulted to the jet not being able to go further forward towards the specimen thus dispersing upwards or downwards according to its location away from its center point where the resistance is less due to the turbulence flow in the impacted media (Finnermore and Franzini, 2002). The turbulent flow is a result of the jetting pushing aside the impacted media away from obstructing it from moving forward.

Proof of Tunnelling Effect during Impact

Tunnelling effect could be describe as a pathway created due to the dispersion of sand to give way for water jet which has a high velocity. During the experiment two types of tunnelling effect was identified. First being the horizontal tunnelling and vertical tunnelling being the later.

Figure 5.5 shows that the existence of the vertical tunnelling with an upwards movement of the jet after it has been impacted upon the pipe surface. Following the pipe curvature the


upwards flow is a lot evident compared to the downward flow because compacted sand resistance on the upper side of the pipe is a lot lower compared to the one beneath it. Compacted sand above the pipe can be easily dispersed due to the condition above the pipe being in the manner of an open system which has direct contact with the atmosphere. Upward movement of the jet resulted to the gradual collapsing of the sand structure above the pipe specimen. It will bring upon a hollow like vertical tunnel which allows the movement upwards steadily without much resistance.

Figure 5.5 Tunneling profiling

Turbulent flow area Compacted

sand

Pipesurface

Verticaltunneling

Horizontal tunneling


Figure 5.6 Tunneling effect on pipe

Horizontal tunnelling existence can be explained by observing the nature of sand dispersion. When water is jet out of an orifice, it will push aside sand particles near the outlet of the orifice. Due to its high velocity, the water jet will act like a solid thus providing enough force to push away sand particles from blocking its way. Proof of the mixing of sand particles with the water jet can be seen through the identification of the rough surface upon the impacted surface and this is especially true as there are two different areas on the impacted surface.

Relationship between Specimen Thickness and Time

Basically, pipe thickness is seen to be decreasing with the increment of time. From the graph it could be postulated that the rate of thinning in Zone 1 is a lot higher compared to the rest of the

Area 1 (enclosed in red circle)

Area 2 (enclosed in blue circle)


lines regardless of the lines on which the points are situated (Figure 5.7). The explanation for this occurrence is that the centre point of impact coincides with Area 1 which has water as the majority erosion media, thus hampering its thinning rate compared to the points in Zone 1 which coincides with Area 2 of the tunneling created.

Pipe Thickness Vs Time (Line A)

0.00E+001.00E-032.00E-033.00E-034.00E-035.00E-036.00E-037.00E-03

0 100 200 300 400

Time (hour)

Thi

ckne

ss (m

)

Point A1Point A3Point A5

Figure 5.7 Analysis of Line A

Dispersion of Specimen Thinning Effect

From Table 1, it is evident that the highest thinning rate is in the Line E region while the lowest thinning rate is evident in the Line A region. This could be explained by the existence of vertical tunnelling near Line A region and turbulent flow area near Line E region as shown in Figure 5.8.


Table 5.1 Average thinning rate of orifice jetting

Point Average Thinning Rate (m/hr) A1 1.08X10-5 A3 2.87X10-6 A5 1.01X10-6 B2 6.84X10-6 B4 1.60x10-6 C1 1.10X10-5 C3 4.46X10-6 C5 1.41X10-6 D2 1.23X10-5 D4 3.28X10-6 E1 1.21X10-5 E3 5.73X10-6

The vertical tunnelling allows minimum contact between abrasive media with surface in Line A region. The explanation for this is that the mass of the sand particles is way too high for it to move upwards following the vertical tunnelling. While in Line E area movement downwards is being resisted by the compacted sand beneath the pipe specimen. Resulting from this is the circular flow of abrasive material in the water jet nearing Line E region as the water jet’s flow is being resisted by the compacted sand beneath the pipe specimen.


Figure 5.8 Close-up of the rupture point (blue enclosure)

Besides the effect of turbulent flow in Line E region too is affected by the gravitational force where the sand with high mass will never be able to travel upwards without being pulled downwards by the gravitational force thus it will stays near Line E region which is a lot easier for it to comprehend rather than moving upwards which requires higher acceleration thus creating a higher concentration of abrasive material in that area.

Zero Effect for 5 mm Orifice Jetting in Water to Sand System

From Figure 5.10, it could be identified that the abrasive effect of the jetting stops completely at distance of 70 cm away from the jetting source. According to the trend of the graph the distance that the abrasive effect upon the impacted surface stops starts at around the distance of 45cm away from the jetting source.


Thinning Rate Versus Distance

3.59E-04

0.00E+002.58E-06

2.36E-05 2.16E-05

-5.00E-050.00E+005.00E-051.00E-041.50E-042.00E-042.50E-043.00E-043.50E-044.00E-04

0 10 20 30 40 50 60 70 80Distance (cm)

Thi

nnin

g R

ate

(m/h

r

Figure 5.10 Thinning rate versus distance

CONCLUSION

Abrasion effect does contribute to the erosion rate of natural gas piping in terms of metal thinning. An increase in exposure time resulted to the decrease of specimen thickness. The existence of two distinct patterns on the impacted surface shows the effect of abrasive material dispersion is closely related to the velocity of the jet in its centerline region and region further away from its centerline. Pattern of impact and the dispersion of metal thinning showed the existence of vertical and horizontal tunneling due to velocity differences and gravitational effect. Besides that the experiment too was able to demonstrate the zero effect distance away from the jetting source to be at 45cm.


REFERENCES

Brunton, J. H.(1961). Deformation of Solid by Impact of Liquids at High Speeds. Symposium on Erosion and Cavitation, 6, 83-98.

DeCorso, S. M. and Kothman, R. E.(1961). Erosion by Liquid Impact. Symposium on Erosion and Cavitation, 3, 32-45.

Finnemore, E.J. and Franzini J.B. (2002). Fluid Mechanics. 10th

ed. New York: McGraw-Hill Higher Education. Finnie, I.(1961). Erosion by Solid Particles in a Fluid. Symposium

on Erosion and Cavitation, 5, 70-82. Greitzer, E.M., Tan, C.S. and Graf, M.B. (2004).Internal Flow,

Concepts and Applications. United Kingdom: Cambridge University Press.

National Transportation Safety Board (2000). Pipeline Incident Report, Natural Gas Pipeline Rupture and Fire, Near Carlsbad, New Mexico, August 19 2000. Virginia: National Technical Information Service.

National Transportation Safety Board (1998). Pipeline Incident Report, Natural Gas Pipeline Rupture and Fire in South Riding, Virginia, July 7 1998. Virginia: National Technical Information Service.

6VAPOR LIQUID EQUILIBRIUM

BEHAVIOUR OF LIQUEFIED PETROLEUM GAS IN STORAGE

Zainal Zakaria S.Y. Tee

INTRODUCTION

The accurate prediction of phase equilibrium of fluid mixtures is extremely important in many industrial applications, such as reservoir modeling, process design, and gas processing and separation. One practicable example of phase equilibrium is the primary process in an oil refinery which involves the separation of the crude oil into the more valuable fractions i.e., gasoline, kerosene, diesel fuel, etc. by distillation. Equilibrium is a static condition in which no changes occur in the properties of a system with time. A state of equilibrium is a state of rest (Lewis and Randall, 1961). Phase-equilibrium thermodynamics seeks to establish the relations among the various properties, in particular, temperature, pressure and composition, that ultimately prevail when two or more phases reach a state of equilibrium wherein all tendencies for changes have ceased.

Most of the initial work in vapor liquid equilibrium (VLE) behaviour of hydrocarbon mixture was with the system at low pressure and low temperature where an ideal state was usually assumed. Based on the previous researches, there are no studies carried out on the hydrocarbon system, especially the light hydrocarbon system, at non-ideal condition. Ideal model like


Raoult’s law model is applied to the hydrocarbon system at ideal state. Situations change when the said system is not at ideal state since ideal systems hardly exist in real life. The deviations from mixture ideality should be accounted in the prediction of VLE for propane butane mixture.

The fugacity is a quantity that corresponds to the pressure for a non-ideal gas. Fugacity is a pseudo or effective pressure. It is the pressure at which the chemical potential of an ideal gas is the same as that of the real gas at the true pressure. Fugacity of a component in a gas mixture is a pseudo or effective partial pressure for that component.

Fugacity if is a property of a pure material and it depends upon temperature and pressure, which must be uniform throughout both phases at equilibrium. The criterion of vapour liquid equilibrium for multicomponent system is as follows:

sati

Vi

sati

Li PTfPTf ,, (6.1)

Fugacity coefficient is another new property, which is dimensionless. The fugacity coefficient of pure species i , i is defined as:

Pfi

i (6.2)

When dealing with ideal gas, 1i and Pf igi . On the other

hand, the definition of the fugacity of a species in solution is parallel to the definition of the pure-species fugacity. Fugacity coefficient of species i in solution is expressed as:

Pyf

i

ii

ˆˆ (6.3)


In contrast to fugacity, activity coefficient is inherently a multicomponent concept that is useful only for mixtures. It is introduced into Raoult’s law to account for liquid-phase non-idealities. In non-ideal mixtures, activity coefficients depend strongly on liquid-phase composition. Ideal solution serves as a standard to which real-solution behaviour can be compared. Activity coefficient is defined in the following expression:

ii

ii fx

f̂ (6.4)

Modified Raoult’s law includes the activity coefficient to account for liquid-phase non-idealities, but it is limited by the assumption of vapour-phase ideality. This can be overcome by introducing the vapour-phase fugacity coefficient. For species i in vapour mixture and in liquid solution, fugacity of species i invapour phase and in liquid phase can be represented by:

Pyf iiv

iˆˆ and iii

li fxf̂ (6.5)

The criterion for phase equilibrium is that these be equal:

iiiii fxPy ˆ (6.6)

In order to calculate with confidence the fugacities in a gas mixture, it is advantageous to use an equation of state where the parameters have physical significance, i.e. where the parameters can be related to intermolecular forces. One equation of state that has this desirable ability is the virial equation of state. The fundamental advantage of the virial equation is that it directly relates fugacities in mixtures to intermolecular forces (Prausnitz et al., 1999). Vapour phase non-idealities in the calculation of


thermodynamics properties near atmospheric pressure and often up to about 1.5MPa can be represented by the virial equation of state with the inclusion of the second virial coefficient only (Virendra et al., 1995). The generalized virial equation has been widely used because it only requires the substance-dependent critical parameters and acentric factor. Generalized virial equation is of greater applicability to all gases. The most important advantage of the virial equation of state for application to phase equilibrium is its direct extension to mixtures (Prausnitz et al., 1999). Mixing rules should be included when dealing with mixture. For two-term truncated form, mixture second virial coefficient is a function of temperature only.

MATHEMATICAL MODELING

This research involves mathematical modeling. Mathcad is used to numerically solve fugacity, fugacity coefficient and activity coefficient for propane butane mixture. All the required inputs like properties of propane and n-butane should be defined in the early stage. Then suitable equations are listed in correct sequences in order to get the final results. There are four parameters to be solved in this study. They are fugacity, fugacity coefficients, activity coefficients and vapour phase compositions for each species in propane butane mixture. Liquid phase composition and system temperature are set before solving the above parameters. In order to establish a mathematical model, appropriate assumptions are made.

The generalized virial equation of state method is suitable for propane butane mixture. That is, the operating condition of propane butane mixture at 10 bars is assumed as moderate pressure while applying this method. Limitation of this method is its applicability to low and moderate pressure. In this study, 10 bars is


considered high pressure for propane butane system but in the other way round it is considered as moderate pressure when applying this method. Next, the required liquid composition of propane butane mixture is referred to the composition at equilibrium state which is obtained through the composition analyzer. Besides, butane in the mixture is actually consisting of n-butane and isobutane. In this project, butane is referred to a mixture of 50% of n-butane and 50% of isobutane. In addition, propane and butane are chemically similar species and both species are non-reactive in a mixture. Therefore, an assumption of fugacity coefficient in gas phase ( vˆ ) equals to the fugacity coefficient in liquid phase ( Lˆ ) is made. This assumption is based on the fact that the pure species fugacity coefficient in gas phase ( v ) equals to the pure species fugacity coefficient in liquid phase ( L ).


This study focuses on the effect of temperature and pressure on the vapour liquid equilibrium behaviour of propane butane mixture. Mathematical modeling has been developed in order to predict the VLE behaviour. The model currently proposed can be used for the prediction of VLE of propane butane mixture in the full composition range.

Effect of Composition

A typical Pxy diagram for six different temperatures can be seen in Figure 6.1. It clearly shows that when there is an increase in temperature, there is an increase in system pressure. For each


temperature, the upper curve represents bubble point line while the lower curve represents dew point line. A temperature range of 263.15 K to 313.15 K chosen in this modeling is of practicability for propane butane mixture in cylindrical storage.

0.000.501.001.502.002.503.003.504.004.505.005.506.006.507.007.508.008.509.009.50

10.0010.5011.0011.5012.0012.5013.0013.5014.0014.50

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1

mole fraction of propane, xC3 or yC3

P (b

ar)

P vs. x at 263.15KP vs. y at 263.15KP vs. x at 273.15KP vs. y at 273.15KP vs. x at 283.15KP vs. y at 283.15KP vs. x at 293.15KP vs. y at 293.15KP vs. x at 303.15KP vs. y at 303.15KP vs. x at 313.15KP vs. y at 313.15K

Figure 6.1 Pxy diagram for several temperatures

Effect of Temperature

Figure 6.2 shows the effect of temperature changes on the fugacity coefficient. Here, it is assumed that mole fractions in liquid phase for both the propane and butane are 0.6 and 0.4, respectively. The same assumption also goes to Figures 3, 4 and 6. It can be seen clearly that fugacity coefficients of solution as well as of individual species in mixture decreases steadily as the system temperature


increases. That is, the deviation of fugacity coefficient from unity becomes larger as the system temperature becomes higher.

0.7900.7950.8000.8050.8100.8150.8200.8250.8300.8350.8400.8450.8500.8550.8600.8650.8700.8750.8800.8850.8900.8950.9000.9050.9100.9150.9200.9250.9300.9350.9400.9450.950

260 262 264 266 268 270 272 274 276 278 280 282 284 286 288 290 292 294 296 298 300 302 304 306 308 310 312 314 316T (K)

fuga

city

coe

ffici

ent

propane : fugacitycoefficient in mixturebutane : fugacity coefficientin mixturesolution fugacity coefficient

Figure 6.2 Fugacity coefficients at different temperatures

Figure 6.3 shows the effect of temperature changes on fugacity as well as system pressure. As shown in Figure 3, solution fugacity increases as the system temperature increases. It is showed that the solution fugacity appears very closely to the fugacity of propane in mixture.

This is because fugacity is known as a parameter representing the effective pressure in vapour phase (Prausnitz et al., 1999). Since there is more propane in vapour phase, therefore both the solution fugacity and fugacity of propane in mixture show the similar trend. That is, there is negligible gap between these two parameters. On the other hand, Figure 3 shows clearly that the system pressure always greater than the solution fugacity. A system is considered as an ideal system when these two parameters


show negligible difference. As shown in Figure 6.3, difference between system pressure and solution fugacity becomes larger when system temperature becomes higher. That is, deviation from ideality for propane butane mixture is more apparent at higher temperature.

0200400600800

100012001400160018002000220024002600280030003200340036003800400042004400460048005000520054005600580060006200640066006800700072007400760078008000

260 262 264 266 268 270 272 274 276 278 280 282 284 286 288 290 292 294 296 298 300 302 304 306 308 310 312 314 316

T (K)

fuga

city

or

pres

sure

(mm

Hg)

propane : fugacity in mixturebutane : fugacity in mixturesolution fugacityvapour pressure

Figure 6.3 Fugacity and vapour pressure At different temperatures

As shown in Figure 6.4, the activity coefficient of butane increases appreciably as temperature grows higher. Meanwhile, there are only insignificant decreases in activity coefficient of propane with temperature. As can be seen, activity coefficient of propane stays closely to unity while the activity coefficient of butane deviates from unity. Activity coefficient is a parameter used to account for liquid-phase non-idealities. In general, there is more butane in liquid phase. Therefore, the effect of temperature on activity coefficient of propane in liquid solution is quite


insignificant while the effect of temperature brings relatively significant changes on butane.

0.9975

0.9996

1.0017

1.0038

1.0059

1.008

1.0101

1.0122

260 262 264 266 268 270 272 274 276 278 280 282 284 286 288 290 292 294 296 298 300 302 304 306 308 310 312 314 316

T (K)

activ

ity c

oeffi

cien

t

propane : activitycoefficientbutane : activitycoefficient

Figure 6.4 Activity coefficients at different temperatures

Figure 6.5 shows the effect of temperature on vapour phase composition. As temperature becomes higher, more butane vaporizes therefore the vapour phase composition of butane increases while the vapour phase composition of propane decreases. However, there is always more propane in vapour phase due to its higher vapour pressure. These vapour phase compositions are then used in the study of discharging process of propane butane mixture from cylindrical storage.


0.000.040.080.120.160.200.240.280.320.360.400.440.480.520.560.600.640.680.720.760.800.840.880.920.961.00

260 265 270 275 280 285 290 295 300 305 310 315 320

T (K)

vapo

ur p

hase

mol

e fr

actio

n

propane : vapour phase composition

butane : vapour phase composition

Figure 6.5 Vapour phase compositions at different temperatures

Effect of Pressure

Figure 6.6 displays the effect of pressure on activity coefficient at different temperatures. At constant temperature, as pressure goes higher deviations of activity coefficient for butane become more noticeable. Even though there is a deviation from unity, the said deviation can actually be neglected because, as can be seen from Figure 6, the highest value of activity coefficient is less than 1.02 for these six temperatures. This value is actually not far from unity. Sometimes, this value can even be approximated to unity. Unless near the critical region, activity coefficient is little affected by pressure and is strongly affected by the nature of pure chemicals comprising the liquid solution (Alvarado, 1993). This statement clearly explains that it is reasonable to neglect the deviation of activity coefficient when pressure changes. Meanwhile, activity coefficient of propane approaches unity at certain pressure. Apart


from that particular pressure, it is clearly shown that activity coefficient goes larger than unity at lower pressure and smaller than unity at higher pressure. Besides, larger range of value of activity coefficient is accounted as temperature increases.

0.995

1.000

1.005

1.010

1.015

1.020

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

P (mmHg)

Act

ivity

coe

ffici

ent

propane : T = 263.15Kbutane : T = 263.15Kpropane : T = 273.15Kbutane : T = 273.15Kpropane : T = 283.15Kbutane : T = 283.15Kpropane : T = 293.15Kbutane : T = 293.15Kpropane : T = 303.15Kbutane : T = 303.15Kpropane : T = 313.15Kbutane : T = 313.15K

Figure 6.6 Effect of pressure on activity coefficient at different solution fugacity at temperatures

Figure 6.7 depicts the effect of system pressure on solution fugacity at different temperatures. It is clearly shown that as the pressure increases, solution fugacity increases at constant temperature. At higher temperature, solution fugacity becomes larger and it obviously accounts for larger range of values. Figure 7 clearly tells that solution fugacity is always smaller than system pressure due to the non-idealities in propane butane mixture. However, the difference between these two parameters for propane


butane mixture is always small. Judging from the modeling, the ratio between solution fugacity and system pressure (f/P) always ranges from 0.73 to 0.87.

650

1050

1450

1850

2250

2650

3050

3450

3850

4250

4650

5050

5450

5850

6250

6650

7050

7450

7850

8250

650 1050 1450 1850 2250 2650 3050 3450 3850 4250 4650 5050 5450 5850 6250 6650 7050 7450 7850 8250 8650 9050 9450 9850

P (mmHg)

solu

tion

fuga

city

(mm

Hg)

solution fugacity vs. P at 263.15Ksolution fugacity vs. P at 273.15Ksolution fugacity vs. P at 283.15Ksolution fugacity vs. P at 293.15Ksolution fugacity vs. P at 303.15Ksolution fugacity vs. P at 313.15K

Figure 6.7 Effect of pressure on different temperatures

Figure 6.8 depicts the effect of system pressure on solution fugacity coefficient at different temperatures. It is clearly shown that as the pressure increases, solution fugacity coefficient decreases at constant temperature. At higher temperature, solution fugacity coefficient accounts for larger range of values and the deviations from unity are larger.


0.82

0.83

0.84

0.85

0.86

0.87

0.88

0.89

0.90

0.91

0.92

0.93

0.94

0.95

0.96

0.97

800 1400 2000 2600 3200 3800 4400 5000 5600 6200 6800 7400 8000 8600 9200 9800

P (mmHg)

solu

tion

fuga

city

coe

ffici

ent

T = 263.15KT = 273.15KT = 283.15KT = 293.15KT = 303.15KT = 313.15K

Figure 6.8 Effect of pressure on solution fugacity coefficient at different temperatures

CONCLUSION

This study enabled prediction of vapour liquid equilibrium behaviour of propane-butane mixture at non-ideal state. The proposed model which is based on the generalized virial equation of state and Gamma/Phi formulation can solve numerically the fugacity, the fugacity coefficient, the activity coefficient and the vapour-phase composition of propane-butane mixture. The results obtained have shown that the solution fugacity coefficient decreases steadily as the system temperature and pressure increased. The vapour-phase composition of propane decreases as the system temperature and pressure go higher. Meanwhile, butane’s concentration in vapour phase becomes richer.


REFERENCES

Alvarado, J. F. J. (1993). Studies on Phase Equilibrium Calculations from Equation of State. Texas A & M University. Ph. D. Thesis.

Hall, K. R. and Iglesias-Silva, G. A. (1993). Quadratic Mixing Rules for Equations of State: Origins and Relationships to the Virial Expansion. Fluid Phase Equilibria. 91 (1): 67-76

Hernandez-Garduza, O., Garcia-Sanchez, F., Neau, E. and Rogalski, M. (2000). Equation of State Associated with Activity Coefficient Models to Predict Low and High Pressure Vapor-liquid Equilibria. Chemical Engineering Journal. 79: 87-101

Kleiber, M. and Axmann, J. K. (1998). Evolutionary Algorithms for the Optimization of Modified UNIFAC Parameters. Computers and Chemical Engineering. 23: 63-82

Lewis, G. N., and Randall, M. (1961). Thermodynamics. New York: McGraw-Hill Book Company.

Prausnitz, J. M., Lichtenthaler, R. N. and Azevedo, E. G. (1999). Molecular Thermodynamics of Fluid-Phase Equilibria. 3rd

edition. New Jersey: Prentice Hall.Virendra, U., Rajiah, A. and Prasad, D. H. L. (1995). Dependence

of the Second Virial Coefficient on Temperature. TheChemical Engineering Journal. 56: 73-76.

7ANALYSIS OF HOUSEHOLD ENERGY

DEMANDZainal Zakaria

Wong Song Harn

INTRODUCTION

Residential sector energy consumption accounts for about 45% of overall energy use in developing countries. The largest portion of residential sector energy used goes towards food preparation, which accounts for 85% of the total (Brandon and Lewis, 1999). Exposure to emissions caused by burning fuels is believed to be responsible for a significant share of the global burden of disease. Energy demand can vary significantly due to residential type and life style of the region. The future trend of residential energy demand is one of the key factors of the climate change problem. There are also other problems to be considered, in particular poor indoor air quality. In the future, increased efficiency of energy systems and reduced end-use energy demand will be important in attaining the 6% curtailment of green house gases targeted by the Kyoto Protocol (Smith, 1999). Reduced energy consumption in the residential sector is particularly important, because energy demand in this sector is notably increasing.

Health researchers believe that millions people die annually as a result of indoor air pollution from cooking. Women and children are much more likely to be affected. The central question of this study hinges on the health impacts of cooking fuel. From the point of view of economics, people are too poor to afford


cleaner alternatives. On this view households invest in clean household energy technologies until the benefits of cleaner air would be more than offset by the loss of utility due to the additional cost of clean fuels. An ideal test of the poor but efficient hypothesis would estimate household’s willingness to pay for risk reductions implicit in fuel switching and compare this measure with other estimates of risk-money tradeoffs. If households are managing the risk of illness from indoor smoke in an efficient manner then the willingness to pay for risk reduction should accord with risk reduction expenditures in other domains (Pranab et al., 2002). The energy demand and expenditure patterns for urban and rural households in the city of Miri are to be discussed as a case example of analyzing residential energy consumption across different population groups in a country. Kitchen performance tests and controlled cooking tests (CCT) are to be applied to measure savings in household fuel wood use from partially switching to LPG.

Production and consumption of almost any type of energy have environmental impacts. Harvesting of fuelwood, in particular, contributes to deforestation, soil erosion, and desertification. Use of fuelwood as an energy source can also contribute to the accumulation of Carbon Dioxide, the main greenhouse gas, both because burning fuelwood produces Carbon Dioxide, and because deforestation destroys an important Carbon Dioxide sink. In addition, use of biomass in traditional stoves exposes the users, mainly women and children, to high levels of indoor air pollution. The key determinants of energy demand in the household sector include: prices of fuels and appliances; disposable income of households; availability of fuels and appliances; particular requirements related to each; cultural preferences (Oleg and Ralph, 1999). With increasing disposable income and changes in lifestyles, households tend to move from the cheapest and least convenient fuels, biomass to more convenient and normally more expensive ones, LPG, and eventually to the most convenient and


most expensive types of energy, electricity. There is a strong positive relationship between growth in per capita income and growth in household demand for commercial fuels.

There is also a correlation between the choice of cooking fuels and the value of women's time. Women who enter the formal workforce demand more convenience in their use of household fuels. Urbanization is an important determinant of both the quantity and the type of fuel used in developing countries. In general, urbanization leads to higher levels of household energy consumption, although it is difficult to separate the effects of urbanization from the increases in income levels that generally accompany urbanization. There is also a shift from traditional to commercial fuels. Several factors that contribute to this trend include a decline in access to biomass fuels, inconvenience of transportation and storage, and improvement in availability of commercial fuels in urban areas. Nonetheless, use of traditional fuels in many cities of the developing world remains high among low-income groups (Erol and Yu, 1988).

METHODOLOGY

A total of 100 households were interviewed on their monthly incomes, monthly consumption on energy, and their preferred future cooking fuels. An interview session was conducted with Sarawak Gas Distribution Sdn. Bhd. on the fuel gas supply system, pattern of consumption and all the relevant information. Kitchen Performance Tests and Controlled Cooking Tests were being conducted, to discuss more detail on the energy consumption for cooking.


Interviews

A sample containing 100 households is selected, covering the urban and rural households. Two residential areas in urban area in Miri are chosen to be the surveyed area, those are, Pujut and Piasau. On the other hand, the rural regions to be interviewed are from two villages 2 hours drive from town, Kampung Puyuh, and Kampung Pasir. The areas were selected trying to keep relatively constant variables such as area compound, traditional cooking practices, climatic conditions and access to LPG. Within each village, households to be sampled are belonging to different socio-economic strata. Parameters needed are location, region, population, number of households, members of households socioeconomic groups (incomes), fuel used, fuel consumption, price of fuels. Besides, a few staffs of Sarawak Gas Distribution Sdn. Bhd. are being interviewed regarding the gas supply system and related data.

Kitchen Performance Tests (KPT)

Kitchen Performance Tests and Controlled Cooking Tests (CCT) are to be applied to measure savings in household fuelwood use from partially switching to LPG. Van Engelenburg et al. (1994). Kitchen Performance Tests (KPT) allow the measuring o f actual household consumption over a pre-determined period of time. The tests are to be applied to 10 households in both urban areas and rural areas. Besides fuel consumption, the number of members per household and their gender and age were recorded. Fuels consumption for three meals are recorded and related to the number of households’ members.


Controlled Cooking Test (CCT)

Controlled Cooking Test (CCT) allows the measurement of comparative fuel consumption under controlled. The test is to be performed with a housewife, in her own kitchen. Measurements taken include three typical meals. Then, the Specific Fuel Consumption, SFC is being calculated by the equation:

SFC=FUEL USED (kg)/Food Prepared (kg).

Pollutants Emissions Analysis

By using the CO2 data logger meter, measurements of Carbon Dioxide levels are taken. It is mainly for technicians and other professionals who need to measure Carbon Dioxide levels within spaces such as industrial environments, commercial buildings or residential dwellings where accumulation of combustion gas is possible. The emission of Carbon Dioxide is positively related to the pollution level.


The main household energy supply sources are biomass, LPG, and electricity, with electricity being used largely for water boiling in all sub-sectors. LPG and electricity are the main forms of energy employed in urban households. Rural sector cooking is confined mainly to biomass. The household income, total electricity use, electricity used for cooking and LPG usage are estimated strongly


positively correlated with each other, while kerosene and biomass use are negatively correlated. The strongest correlation will be seen between income and total electricity use, which is followed by the correlation between income and electricity used for cooking.

Cooking Fuels Consumption

Most of the low-income households in rural regions of Miri rely on fuelwoods or charcoal as their primary cooking fuel, while over 90% in the highest-income category use LPG. Besides a single primary cooking fuel, many households make use of one or more secondary fuels for supplemental purposes, as a backup, or for fuel-specific cooking activities.

Figure 7.1 indicates the various fuel consumption patterns for different ethnic group. It shows that electricity is the main cooking fuel for both Malay and Chinese, which is 54.32% for Malay and 63.63% for Chinese respectively. LPG is the secondary cooking fuel for both Malay and Chinese regions. LPG usage of Malay 45.68%, while Chinese, 36.37%. Generally, Chinese depends more on electricity due to their living standard and culture. As for the other races such as Iban and Penan, who live mostly in rural area, use fuelwoods as their domain cooking fuel, 86.26%. It may be due to the supplement and availability of the fuel. Most of the rural areas do not have the electricity supplement and continuous LPG supplement.


Type of Races

Perc

enta

ge o

f Con

sum

ptio

n (%

)

Figure 7.1 Fuel Consumption Pattern for Different Ethnic Group

Figure 7.2 shows the percentage of fuel preferred by different ethnic group. Overall, LPG is the most popular choice, 62%, followed by electricity, 37%, and lastly, fuelwood, 1%. Both Malay and Chinese will prefer electricity than LPG, by the ratio of 4:2 and 32:3 respectively. As for the other races, LPG has the highest rate, by the ratio of 57:1:1 towards electricity and fuelwood.


Type of Races

Perc

ent (

%)

Figure 7.2 Fuel Preferred by Different Ethnic Group

Figure 7.3 indicates the effect of number of household members on fuel consumption is urban area. Commonly, the fuel consumption is higher for the household with more members. We may conclude that when the number of members reaches five, the consumption of LPG drops while the electricity consumption increase. It may happen because electricity is more convenient, faster and safer, comparing to the others.


Figure 7.3 Effect of Household Member on Fuel Consumption in Urban Area

Figure 7.4 indicates the effect of number of household members on fuel consumption is rural area. The usage of LPG is very rare in rural area due to the availability and insufficient supplement. There is no significant effect of household members on LPG consumption. However, fuelwoods consumption is related positively towards the number of household members. Commonly, more household members mean more food to be consumed, and more cooking fuels are needed to have sufficient cooking energy.


Figure 7.4 Effect of Household member on Fuel Consumption in Rural Area

Figure 7.5 shows the difference of fuel consumption pattern for urban and rural region. Electricity is the main fuel in urban area but not in used in rural area. Electricity is not supplied to most of the rural area. Only certain households generate electricity by diesel, and it is only to run certain electric appliances such as fans and lights. Fuelwood is the domain in rural area due to the free resource from the surrounding. Thus, it can be concluded that the consumption pattern for rural region depends mostly on the availability of the supplement of the fuels.


Figure 7.5 Fuel Consumption Pattern in Different Region

Figure 7.6 shows the fuel consumption patterns of households with different range of incomes. Household with the income in the range less than RM 1000 are mostly living in rural area. Fuelwoods is their main cooking fuel, with the percentage as high as 94.74%, remanding 5.26% goes to LPG. As stated above, electricity is not being used as cooking fuel in rural area. Most of the households with the income in the range of RM 1000 to RM 2000 live in rural area, with a few exceptions. Fuelwoods, which is the domain cooking fuel in rural area, is their main fuel that is 52.65%, followed by LPG, 35.97% and lastly, electricity, 11.38%. Both households with the income in the range of RM 2000 to RM 3000 and higher than RM 3000 use electricity as their main cooking fuel, with the percentage of 55.72% and 62.31% respectively. To them, expenses on the electricity are affordable and they are leading a better housing condition. Thus, electricity will be a better choice due to the convenience and effectiveness of it.


Households Monthly Income (RM)

Perc

enta

ge o

f Con

sum

ptio

n (%

)

Figure 7.6 Effect of Household Income on Fuel Consumption

Figure 7.7 shows the combustion of fuelwoods produces the higher rate of Carbon Dioxide as the by product, followed by LPG. The emission of electricity is negligible. However, the Carbon Dioxide emission is much lower than the actual reading, due to the open air condition. Theoretically, the emission of Carbon Dioxide of LPG is much lower than fuel wood. There is no any direct emission of any pollutant from electricity. Thus, electricity is the most environmental friendly cooking fuel.


0

50

100

150

200

250

300

FUELWOOD LPG ELECTRICITY

Type of Fuel

Car

bon

Dio

xide

Em

issio

n (p

pm)

Figure 7.7 Carbon Dioxide Emissions for Different Fuel

Kitchen Performance Tests

In kitchen performance test, fifteen households with five members were selected, divided into three groups using different cooking fuel. During the test, food were cooked by one and the fuel used were fuelwoods, LPG and electricity. Table 1 shows the average of fuel consumption for each group. The lowest expense goes to electricity, followed by LPG and lastly fuelwoods. Thus, we may conclude that electricity is the most efficient cooking fuel with the highest performance, lowest expense.


Table 7.1 Kitchen Performance Tests

FUEL TYPE FUEL CONSUMPTION, RM FUEL WOOD/KEROSENE 31.44

LPG 30 ELECTRICITY 26.4

Controlled Cooking Test

Three households using different cooking fuel were under observation during meals preparation. Fuel consumptions to prepare five kilograms of chicken curry were recorded for each household. Table 2 shows that fuelwoods consumed the most, followed by LPG, lastly electricity. Therefore, we get the same conclusion as per Kitchen Performance Tests, electricity is the most effective cooking fuel, followed by LPG and lastly fuelwoods.

Table 7.2 Controlled Cooking Test

FUEL TYPE FOOD PREPARED,KG

FUEL CONSUMPTION,RM

SFC, RM/KG

FUELWOOD 5 1.6 8 LPG 5 0.7 3.5

ELECTRICITY 5 0.125 0.625


CONCLUSION

LPG and electricity are the main energy employed in urban households. Rural sector is confined mainly to biomass. The energy demand, both fuel type and usage are highly related to the household income and household characteristic, as well as races. Switching to LPG might also be faster in mid-high to high income households; large-scale switching by the mid-low and low income households is unlikely under the current circumstances. Commonly, it is known that biomass produces the most Carbon Monoxide, as the by product of the combustion; followed by Diesel, LPG and lastly electricity. The difference in the lifestyle and culture for different ethnic group definitely produce different fuel consumption pattern. Thus, the fuel consumption pattern is affected by the races of the households. Besides, fuel preferred by various ethnic groups also depends on their background and culture. Commonly, as the number of household members increase, the fuel consumption increase positively. It is also noticed that, with more female members or with more members in the range of age of 15 to 50, the fuel consumption is higher relatively. There is a significant difference of fuel consumption pattern in rural area comparing to urban area. This may due to the supplement and availability of the fuel. Carbon Dioxide emission is taken to represent the pollutant emission of the cooking fuels.

REFERENCES

Brandon G. and Lewis A. (1999). Reducing Household Energy


Consumption: A Qualitative and Quantitative Field Study. Journal of Expected Psychol. 7(19):58–74.

Erol, U. and Yu S.H. (1988). On the Causal Relationship between Energy and Income for Industrialized Countries. Journal of Energyand Development. 13(1): 113-122.

Oleg D. and Ralph C. (1999). Trends in Consumption and Production: Household Energy Consumption, DESA Discussion Paper. 6(1): 283-303.

Pranab B., Jean-Marie B., Sanghamitra D., Dilip M., and Rinki S.(2002). The Environmental Impact of Poverty: Evidence from Firewood Collection in Rural. Nepal: Mimeo.

Smith,K. (1999). Pollution Management: Indoor Air Pollution. Focus Discussion Note World Bank. 4(1):1-12.

8NEWTON LOOP METHOD IN GAS

PIPELINE NETWORK Zulkefli Yaacob

Norhana Mohamed Rashid

INTRODUCTION

Recently, natural gas is widely used as an energy source aside of petroleum oil, electrical source and coal. At Malaysia, natural gas has high demand on power generation (71%), large industry (17%), exportation (7%) and others for commercial and resident (5%) [6]. In fact, there are many advantages of natural gas in terms of safety, environmental friendly, easy in handling and cost saving. Natural gas has controlled volume substance and has low specific gravity that is difficult to ignite. There is also provision for safety devices such as valve in pipeline system. Natural gas is very familiar with environmental friendly because it produces clean combustion, lower carbon dioxide as well as lower sulphur oxide and nitrogen oxide emission. The cost of using natural gas is cheaper and produce saving when compared to fuel oil or other petroleum fuel. For a natural gas distribution system, there is a technology called network analysis that has been used to increase the services competency to ensure gas will be supplied to customers continuously at all time. The main consideration in any gas distribution system is to maintain the highest safety standards. This is carried out by ensuring that preventable accidents are avoided at all costs and problems with the distribution network are remedied in a timely fashion.


Practices in network analysis vary widely, especially for non-computer solution, since individually companies simplify calculations by using approximations applicable to their individual situations. The trend toward computer solutions since the great numbers of calculations which involve in complex distribution analysis are rapidly performed checked and conveniently stored within the computer memory system. Simpler network problems may be solved with manual calculators. Organizations that cannot justify investment in computers may use computer service facilities or participate in a group facility. Nowadays, there are several method was used to solve network analysis calculations such as Hardy Cross method, Newton Nodal method, Newton Loop method and Newton Loop Node method. In gas piping system, it is important to make sure that the pressure within the loop is under maximum allowable operating pressure. As a gas supplier, customer will be the main priority, as to ensure continuous supply of gas to them. However, to vary the load of gas to customer, pressure is the main parameter to be considered, to avoid failure in the pipeline. A systematic network analysis is an easy technique that can be applied to gas distribution system especially for a complex calculation of gas pipeline network which is normally done by simulators. In many softwares available at present time, non make comparison on what numerical method used in Newton Loop Method that will give the fastest solution in the simulation process.

METHODOLOGY

In this study, a pipeline route should be identified in order to determine the pipeline network. Here, this route is already specified for each case in terms of flow direction and initial

Newton Loop Method in Gas Pipeline Network 147direction of loop flow. There are three major cases involve in this study, which is two-loop, three-loop and four loop system. Network analysis is carried out by applying Newton Loop Method. The flow equation used in the network analysis is Lacey equation which is for low pressure system. The condition of gas flow in pipeline is assumed to be steady state.


Pipeline Routes and Looping Diagram

The pipeline system has been specified for two-loop, three-loop and also four looping systems. The direction for each pipe involves and initial loop flow is assumed for the first guess. This flow direction will be corrected within the calculation and it’s iteration. Figure 8.1 shows the diagram of two looping system, Figure 8.2 is for three looping system and Figure 8.3 for four looping system.

N3

S1

Q2Q1 Q3

N2

L2 L3 L4

N4

L1

Q4 Q5

Loop A Loop B

Figure 8.1 Gas pipeline network for two looping system


From Figure 8.1, the direction of gas flow which is pipe one, pipe two and pipe three from supply source at S1 is away from the source. For pipe four and pipe five, the direction are from node three to node two and from node three to node four. Direction of loop flow for loop A is counter clockwise and loop B is clockwise. This loop flow direction is assumed and fixed in the calculation. But for gas flow in each pipe, the direction can be changed due to flow rates and pressures in the pipe.

Q

Q

S1

Loop ALoop BLoop C

Q

L5

Q

L3L4 L2

Figure 8.2 Gas Pipeline Network for Three Looping System

Newton Loop Method in Gas Pipeline Network 149

Figure 8.3 Gas Pipeline Network for Four Looping System

When the gas flow in a looping system has been determine, the flow rate of gas through the pipe is assumed to be equal with load out from each node. Meanwhile, flow rate at pipe four and five is set to be zero for initial calculation. For initial value of loop flow need to be set and usually is equal to zero. Then, calculate the value of constant, k in the equation. This constant will be used to determine the pressure drop in each pipe. For each loop involves, determine the loop error equation by considering the flow direction. The tolerance of the iteration should be specified, where in this study, it is equal to 0.05%. If the error is less than the tolerance, the solution process will stop. However, if the error is more than tolerance set, then the solution process will proceed to the next step. Then, calculate the Jacobi matrix and solve the matrix equation by using Gauss Elimination method to determine the correction loop flow for each loop. When correction loop flow

LLLLL

Loop A

Loop B

Loop C

Loop D

S

QQ Q1Q

Q


is obtained, second iteration can be performed by repeating the step to calculate the branch flow with new loop flow until the step to calculate the new loop flow. Iteration can be stop when an error is less than the tolerance setting. This simulation method can be simplified as Figure 8.4.

Figure 8.4 Newton loop method flowchart

Iteration

Calculate initial approximations to branch flows Q0

Set initial value of loop flow q0

Calculate branch flow Qk

Are all nodal error less ten specified

tolerance?

Solution obtained - STOP

Calculate loop Jacobi matrix Jk and solving set of equations

Jk( q)k-[F(q)]k

Calculate new loop flow qk

Iteration K=k+1

Calculate loop error F(q)k

Ye N

Newton Loop Method in Gas Pipeline Network 151In Newton Loop method simulation programming, the

calculation step will be set as follow. The initial value will be used and is based on the Figure 8.4. The calculation step is as follow:

1. An initial approximation to the branch flow is assumed that the loads at the nodes are supplied via the tree branches only (dendrite branch), the flow in the chords are equal to zero.

2. Calculate the constant value for each pipe. The equation is from Lacey’s equation which is given by:

k

kk D

Lk 3107.11

Where k is the number of pipe in the gas pipeline network

3. Calculate the pressure drop and KQ for each branch by using equation:

2kk Qk

Where k is the number of pipe in the gas pipeline network

4. Set the value of loop flow, qo equal to zero for each loop.

5. Start the calculation by completing the requirement in table in Figure 8.5. This calculation can be done by manually and also by using Microsoft Excel. In this study, the calculation has done by using Microsoft Excel 2003.

6. Calculate the loop error for each loop involve. The equation used is based on:


3.75336

-0.69814

532

421

BLoop

ALoop

7. Identify the jacobi matrix to do the correction loop flow.

0.07010.05850.058530.08469259

2J

8. Find the correction loop flow for the loop system.

7534.36981.0

1403.01170.01170.01694.0

B

A

qq

9. Solve the correction loop flow by using Gauss Elimination method. The result is given by:

Forward elimination step:

2356.46981.0

0594.001170.01694.0

B

A

qq

Back substitution step:

2102.713089.53

B

A

qq

Newton Loop Method in Gas Pipeline Network 15310. Calculate the new loop flow by:

2102.71

3089.53

Bo

B

Ao

A

qqq

qqq

11. Calculate new value of flow rate, Q. The example of calculation is given for first flow rate. Same method is used for each new flow rate.

6911.1693089.532501Q

12. Proceeds with the second iteration until the value of pressure drop approximate zero or convergence. Figure 8.6, Figure 8.7 and Figure 8.8 are for the next iteration.

After running the simulation, the value of flow rate and pressure at each pipe and node are obtained and given in Figure 8.5.


Figure 8.5 Gas pipeline network with complete flow rate

The same method will be used to calculate the flow rate and pressure for three-loop and four-loop gas network system. The equation use to get the flow rate and pressure in Newton Loop method will be programmed in the FORTRAN simulator. The developed Newton Loop Method simulator for gas pipeline network system use two different numerical method approach which are Gauss Elimination and Gauss Siedel. This simulation successfully works for three types of looping system which are two-loop, three-loop and four-loop systems. The iteration time is counted by using setting time in the simulator to examine the effect of increasing in flow and number of loop in all systems. In this research, there are four case study that have been used. Different conditions have been set for each case study. The

N3

S1

85.04 217.71 227.29

N2

250 100 180

N4

530

32.34 47.295

Loop A Loop B

Newton Loop Method in Gas Pipeline Network 155case study was divided based on different numerical technique used. Each numerical technique is applied to different loop number. Case study one will use Gauss Elimination and Gauss Siedel technique for two looping system to determine flow rate and pressure. For case study two and three, numerical technique and parameter obtained is the same with the first case but the number of loop is increased to three for case study two and four for case study three. For case study four, the different tolerance will be used at different numerical technique and number of loop. The iteration number will be obtained from here and the accuracy can be determined at different tolerance level. And the last case study is to obtained computation time at different number of loop and numerical technique used. The detailed of all the case study are categorized in Table 8.1.

Table 8.1: Case study category

Case Study

Number of Loop

Numerical Method Used

Parameter Obtained

1 2 Gauss Elimination Flowrate Gauss Siedel Pressure 2 3 Gauss Elimination Flowrate Gauss Siedel Pressure 3 4 Gauss Elimination Flowrate Gauss Siedel Pressure

4 2,3 and 4 Gauss Elimination Iteration number

at Gauss Siedel different tolerance 5 2,3 and 4 Gauss Elimination Time at different Gauss Siedel numerical method


Case Study One

Case study one considers two loop gas network systems with total flow rate supply equal to 530 m3/hr at pressure of 30 mbar and specific gravity of 0.589. The two-loop gas network system is shows in Figure 8.10. The direction of gas flow in the pipe and also the flow rate through the pipe is assumed for initial calculation. The nodal data and pipe data of gas network for two looping system is shown in Table 8.2 and Table 8.3 respectively.

Figure 8.6 Schematic diagram of gas network for case one

N3

S1

Q2 Q1 Q3

N2

L2 L3 L4

N4

L1

Q4 Q5

Loop A Loop B


Table 8.2: Nodal data and properties for case study one

Node Status Pressure (mbarg) Load (m3/hr)

1 Supply source 30 -530

2 Load - 250

3 Load - 100

4 Load - 180

Table 8.3: Pipe data and properties for case study one

PipeSending

node Receiving node Diameter

(mm) Length (m)

1 1 2 150 680

2 1 3 100 500

3 1 4 150 420

4 3 2 100 600

5 3 4 100 340

Case Study Two

Gas network system in this case study is extended from two-loop to three-loop systems. The condition and equation used in this case study is the same as in previous case study. The supply load used is 700 m3/hr. The Nodal data and pipe data is given in the Table 8.4 and Table 8.5 respectively. Figure 8.7 shows the schematic diagram for case study two.


Table 8.4: Nodal data for case study two

Node Status Pressure (mbarg) Load (m3/h)1 Supply source 30 -700 2 Load - 250 3 Load - 150 4 Load - 100 5 Load - 200

Table 8.5: Pipe data for case study two

Pipe Sending


(mm)Length

(m)1 1 2 150 680 2 1 3 100 500 3 1 4 150 420 4 1 5 100 550 5 3 2 100 600 6 4 3 150 400 7 4 5 100 350


Figure 8.7 Schematic diagram of gas network for case study two

Case Study Three

In case study three, another one loop is added to the gas network system. The numerical techniques used are still the same which are Gauss Elimination and Gauss Siedel. Similar pressure supply and specific gravity are used. Load supply of 750 m3/h for this case study was used. Detailed nodal and pipe data is given in table 8.6 and 8.7 respectively. The diagram for gas network for this case study is shown in Figure 8.8.

Q

Q

LoopLoopLoop

Q

L

Q

LL L

S


Table 8.6: Nodal data for case study three

Node Status Pressure (mbarg) Load (m3/hr)

1 Supply source 30 -750

2 Load - 250

3 Load - 150

4 Load - 100

5 Load - 150

6 Load - 100

Table 8.7: Pipe data for case study three

Pipe Sending


(mm)Length

(m)

1 1 2 150 680

2 1 3 100 500

3 1 4 150 420

4 1 5 100 550

5 1 6 150 500

6 3 2 100 600

7 3 4 150 400

8 5 4 150 500

9 5 6 100 350


Figure 8.8 Schematic diagram for case study three

Case Study Four

Case study four is done for accuracy analysis applied to the Newton Loop Method while changing the tolerance level for each two-loop, three-loop and four-loop gas network system. There are three different tolerance that will be specified which are at 5%, 0.05% and 0.0005%. From this analysis, the iteration number will be obtained, which subsequently gives computation time. The detailed of the case study four properties are shown in Table 8.8.

LLLLL

LooLooLooLoo

S

QQ QQ

Q


Table 8.8: Properties for case study four

Tolerance Number of loop Numerical method used

5% 2,3 and 4 Gauss Elimination

Gauss Siedel

0.05% 2,3 and 4 Gauss Elimination

Gauss Siedel

0.005% 2,3 and 4 Gauss Elimination

Gauss Siedel

Case Study Five

Subsequent analysis is done to compare the computation time between numerical technique used which is Gauss Elimination and Gauss Siedel. This comparison will be applied to two, three and four loop gas network system. The computation time to solve the problem will be taken for three times and the average value will be calculated.

8.3 New Flowrate and New Pressure

The study to get the new value of flow rate and pressure for the network system by using Newton Loop method is done in order to achieve the first objective of the study. This study is done for two-loop, three-loop and four-loop gas network system. The result is obtained for case study one for two loops, case study two for

Newton Loop Method in Gas Pipeline Network 163three loops and case study three for four looping system. The results of the new flow rate and new pressure for these three cases are shown in Table 8.9 and Table 8.10.

Table 8.9: New flow rate for case study one, case study two and case study three

No. Loop Numerical method type Pipe New flowrate (M3/hr)

2 Gauss Elimination Q1 217.6816 Q2 85.0475 Q3 227.2708 Q4 32.3183 Q5 -47.2708 2 Gauss Siedel Q1 217.7088 Q2 85.0366 Q3 227.2546 Q4 32.2912 Q5 -47.2546 3 Gauss Elimination Q1 227.5213 Q2 92.9792 Q3 268.5784 Q4 110.9211 Q5 22.4786 Q6 -79.4994 Q7 89.0789 3 Gauss Siedel Q1 227.5213 Q2 92.9792 Q3 268.5784 Q4 110.9211 Q5 22.4786 Q6 -79.4994 Q7 89.0789


Table 8.9: New flow rate for case study one, case study two and case study three

No. Loop Numerical method type Pipe New flowrate (M3/hr)

4 Gauss Elimination Q1 199.4547 Q2 63.7103 Q3 137.3405 Q4 28.8682 Q5 84.2967 Q6 50.5453 Q7 -136.8351 Q8 99.4945 Q9 15.7032 4 Gauss Siedel Q1 199.5216 Q2 63.8112 Q3 137.9202 Q4 27.7273 Q5 85.6054 Q6 50.4784 Q7 -136.6672 Q8 98.747 Q9 14.3945


Table 8.10: New pressure for case study one, case study two and case study three

No. Loop Numerical method type Node New pressure (mbarg)

2 Gauss Elimination S1 30 N2 25.0354 N3 25.7686 N4 26.6575 2 Gauss Siedel S1 30 N2 25.0342 N3 25.7697 N4 26.658 3 Gauss Elimination S1 30 N2 24.5764 N3 25.3321 N4 22.0827 N5 29.6453 3 Gauss Siedel S1 30 N2 24.5764 N3 25.3321 N4 22.0827 N5 29.6453


Table 8.10: New pressure for case study one, case study two and case study three

No. Loop Numerical method type Node New pressure (mbarg)

3 Gauss Elimination S1 30 N2 25.832 N3 27.6255 N4 28.7794 N5 29.4637 N6 29.4526 3 Gauss Siedel S1 30 N2 25.8292 N3 27.6179 N4 28.7691 N5 29.5052 N6 29.4354


Graph: Deviation of flowrate between manual calculation and simulation result at pipe Q1

217.6816

227.416

199.5057199.4547

227.4159217.6396

180

190

200

210

220

230

2 3 4Numbers of loop

Flow

rate

Q1

(m3/

h)

Simulation Manual Calculation

Figure 8.9 Comparison of flow rate between manual calculation and simulation at pipe Q1

According to the graph in Figure 5.4 above, there are very small deviation between manual calculation and simulation by using Newton Loop method programming. The error for this deviation is less than 0.05 % for all three cases. For the two-loop system, the flow rate at pipe Q1 from manual calculation and simulation are 217.6396 m3/h and 217.6816 m3/h respectively. The error for the two-loop system is 0.019 %. For the three-loop system, the flow rate from manual calculation and simulation are 227.4159 m3/h and 227.416 m3/h respectively. The error for three-loop system is 0.0000439 %. For the four looping system, manual calculation and simulation 199.5057 m3/h and 199.4547 m3/h respectively. The error is about 0.025 %. From the result obtained, Newton Loop method’s programming can be used to calculate the flow rate in gas pipeline network system with smaller error if compared to manual calculation.


Graph: Deviation of pressure between manual calculation and simulation result at node 2

25.0354

24.5814

25.832 25.8299

24.581525.0373

23.5

24

24.5

25

25.5

26

2 3 4Numbers of loop

Pre

ssur

e (m

barg

)

Simulation Manual calculation

Figure 8.10 Comparison of pressure between manual calculation and simulation result at node two

From manual calculation, the pressure for the two-loop system is 25.0373 mbarg and 25.0354 mbarg is given by simulation and the error is 0.0076 %. Three-loop system also gives small error which is 0.0000813 % with pressure from manual calculation equal to 24.5815 mbarg and 24.5814 mbarg by simulation. For four-loop system, the pressure from manual calculation is 25.8299 mbarg and 15.832 mbarg from simulation. The error is about 0.0081 %. The overall error for pressure calculation by using this programming compared to manual calculation is less than 0.05 %. Thus, this programming can be used to calculate pressure in gas pipeline network system with relatively small error.

Newton Loop Method in Gas Pipeline Network 169Accuracy Analysis of Different Tolerance Used

In this program, different tolerance levels are used to get more accurate answer for flow rate and pressure for the system. High tolerance and low tolerance that are required can be set in the programming. The simulator can perform the calculation and stop at the specified tolerance. The effect of different tolerance used is shows in Table 8.11.


Table 8.11: Different tolerance set affect accuracy of the answer

Tolerance No. of Loop Numerical Type Iteration No.

5% 2 Gauss Elimination 3

Gauss Siedel 3

3 Gauss Elimination 3

Gauss Siedel 3


Gauss Siedel 3

0.05% 2 Gauss Elimination 4

Gauss Siedel 7


Gauss Siedel 6


Gauss Siedel 9

0.0005% 2 Gauss Elimination 5

Gauss Siedel 11


Gauss Siedel 10


Gauss Siedel 16

There are effects of different tolerance set in the program to solve the gas network problem. From graphs in Figure 8.11, Figure 8.12 and Figure 8.13, higher tolerance will results in more iteration number for the calculation. For two-loop system, at 5 % tolerance the iteration number is equal to three. But when reducing the

Newton Loop Method in Gas Pipeline Network 171tolerance level to 0.05 %, the iteration number is also reduce to four by using Gauss Elimination and six for Gauss Siedel. After reducing to 0.0005 % tolerance, the iteration numbers become five for Gauss Elimination and eleven for Gauss Siedel. It shows that, the smaller tolerance specified will create more iteration numbers and will give higher accuracy to the flow rate and pressure calculated.

Graph: Comparison of iteration numbers at different tolerance for two loop system

34

5

3

6

11

0

2

4

6

8

10

12

5% 0.05% 0.0005%Tolerance (%)

Itera

tion

Num

bers

Gauss Elimination Gauss Siedel

Figure 8.11 Effect of tolerance used on iteration numbers at two loop system

According to the graph in Figure 5.7, at low tolerance level which is 5 % the iteration numbers is three. When reducing the tolerance to 0.05 % and 0.0005 %, the iteration numbers increase to six and ten for both numerical techniques used. Low tolerance used was affecting the numbers of iteration and also the accuracy of the programming system. Iteration numbers also indicates the


computation time of this simulation. This is because increasing in iteration numbers will ultimately increase the computation time. Iteration is included in the solving process in the simulation in order to achieve specified tolerance in the system and thus producing more accurate flow rate and pressure. Similar condition applied in a four looping system shown in Figure 8.12 which is the at 5 % tolerance. The iteration number needed is just three but the accuracy is low. At 0.05 % tolerance, Gauss Elimination method gives six iteration numbers while Gauss Siedel gives nine iterations for the calculation. When reducing the tolerance to 0.0005 %, a high iteration numbers is obtained which are twelve for Gauss Elimination and six teen for Gauss Siedel. Here, the effect of tolerance to the iteration number and also to the computation time is increased when a smaller level of tolerance is specified.

Graph: Comparison of iteration numbers at different tolerance for three loop system

3

6

10

3

6

10

0

2

4

6

8

10

12

5% 0.05% 0.0005%Tolerance

Itera

tion

Num

bers


Figure 8.12 Effect of tolerance used on iteration numbers at three loop system


Graph: Comparison of iteration numbers at different tolerance for four loop system

36

12

3

9

16

0

5

10

15

20

5% 0.05% 0.0005%Tolerance

Itera

tion

Num

bers


Figure 8.13 Effect of tolerance used on iteration numbers at four loop system

CONCLUSIONS

As a conclusion, the constant flow rate and pressure can be obtained by using Newton Loop Method simulation program within the loop error stipulated. Pressure drop in the looping system will converge to zero and minimize the loop error. This program also can be done for the various types of gas network system such as two-loop, three-loop and four-loop system. The flow rate for two-loop, three-loop and four-loop system is given in Table 5.9 while the pressure for two-loop, three-loop and four-loop is shown in Table 5.10 respectively. The new value of flow rate and pressure obtained depend on the initial load and pressure that has been supplied to the gas network system and also the tolerance


for the calculation. The specific tolerance has been chosen for the calculation process which is equal to 0.05 %.

REFERENCES

Osiadacz A.J. (1987) Simulation and Analysis of Gas Networks. London: E & F.N. Spon Ltd

Yang, K.W (2001) Convergence Characteristics of Steady State Gas Network System Using Numerical Loop-Node Method. Master Thesis. Universiti Teknologi Malaysia.

Poh,Hong Hwee (2005). Application of Computer Based Simulation Gas Network System Using Graph Theory Algorithms and Numerical Method. Master Thesis. Universiti Teknologi Malaysia.

Cormen Thomas H., (2002). Introduction to Algorithms., 2nd ed. Cambridge..The MIT press

Jaap Van den Herik, Jos Uiterwijk, Jeroen Donkers, Sander bakes, Guillaume Chaslot and Jahn-Takeshi Saito, intelligent Search Techniques., Universiteit maastricht.

Zulkefli Y. Natural Gas Transmission System, Lecture Note, University Technology of Malaysia. Unpublished.

Steven C. Chapra & Raymond P. Canale. Numerical Methods for Engineers, Fifth Edition. Mc Graw-Hill.

Md Shah Majid, Hasimah Abdul rahman, Norzanah Rosmin, Mohd. Hafiz Habibudin, Dalila Mat Said, Rasyidah Mohd. Idris & Saifulnizam Abd. Khalid (2005). Prinsip Kejuruteraan Elektrik, Edisi Pertama. Jabatan Kejuruteraan Power Fakulti Kejuruteraan Elektrik, Universiti Teknologi Malaysia. (1-12 - 1-13)

INDEX

Abrasion, 101, 105, 114 Activity coefficient, 119,

124, 125 Alloy, 75, 76, 77, 78 Brass, 75, 76, 77, 78 Butane, 64 CFD, 14, 16 CNG, 32, 33, 34, 57, 58, 65,

75, 83, 97 CO, 32, 36, 38, 39, 40, 44,

45, 46, 49, 50, 51, 57, 65, 83, 85, 88, 97

CO2, 32, 36, 38, 39, 40, 44, 46, 47, 48, 49, 57, 64, 87, 134

Coating, 103, 105 Computational, 14 Consumption, 57, 97, 134,

135, 136, 138, 139 Dispersion, 108, 111 Dissipation, 10, 17, 21, 22,

23Efficiency, 98 Electricity, 138 Emission, 43, 58, 81, 85, 91,

98ENAREC, 86 Energy demand, 131 Engine, 33, 34, 35, 36, 40,

41, 58, 84, 91, 94, 96, 97, 98

Engine oil, 84, 91, 94

Environmental, 97, 98 Erosion, 100, 101, 115 Exhaust, 33, 43, 57, 85, 97 Finite element, 62 FLUENT, 13, 16, 28 Fluid, 14, 29, 115, 129, 130 Fuel, 31, 57, 58, 87, 97, 134,

136, 137, 138, 139 Fugacity, 118, 123, 124 GAMBIT, 12, 13 Gas, 28, 31, 32, 35, 37, 40,

41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 63, 64, 66, 67, 68, 70, 79, 82, 98, 115, 133

Gas mixer, 31, 32 Gasoline, 58, 65, 83, 87, 92 Gear, 41, 42, 43, 47, 48, 49,

50, 51, 52, 53, 54, 55, 56 Household, 138, 139 Hydrocarbon, 64, 90 Idling, 38, 39, 44 Ignition, 33, 34, 64, 84 Impact, 100, 108, 115 Jetting, 108, 113 Kitchen performance test,

131Liquid phase, 120 Mathcad, 120 Mathematical modeling, 121 Metering, 2, 11, 28

176 Index

Modelling, 29 Motorcycle, 31, 34, 59, 62,

70, 81 Natural gas, 36, 61, 63, 64,

66, 83, 87, 91, 93, 98 NGV, 31, 32, 33, 57, 63, 65,

66, 67, 68, 69, 78, 79, 82, 98

NGVM, 61, 62, 70, 71, 72, 73, 74, 75, 78

non-methane hydrocarbon (NMHC)., 83

NOx, 32, 36, 38, 39, 40, 44, 46, 54, 55, 56, 57, 83, 86, 90, 91, 92, 94

Orifice, 2, 3, 4, 17, 18, 19, 20, 21, 25, 28, 29, 107, 113

PATRAN / NASTRAN, 74 PETRONAS, 82, 87 Pollutants, 134 Population, 97 Pre-mixing, 32 Pressure, 62, 68, 72, 88, 126,

130

Production, 131 Profiling, 108 Propane, 64 Region, 17 Regulator, 68, 69, 70, 71, 72,

77Residential, 130 Rupture, 115 Rural, 135, 139 Safety, 75, 99, 115 Solidified, 72 Standard, 8, 16, 17 Structural analysis, 70 Thinning, 104, 105, 111, 112,

114Thinning rate, 114 Transmission, 34, 35 Tunnelling, 108 Turbulence, 29 UHC, 44, 45, 52, 53, 54, 90,

97Upstream, 6 Vanes, 7, 9 yield, 75

advanced gas tech i toc+preface - home | official web ...€¦ · content preface vii chapter 1...

Documents