- Email: [email protected]
Improving gas transmission networks operation using simulation algorithms: Case study of the National Iranian Gas Network
Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Contents lists available at ScienceDirect
Journal of Natural Gas Science and Engineering journal homepage: www.elsevier.com/locate/jngse
Improving gas transmission networks operation using simulation algorithms: Case study of the National Iranian Gas Network Maryam Fasihizadeh a, Mohsen V. Sefti a, *, Hassan M. Torbati b a b
Department of Chemical Engineering, Tarbiat Modares University, Tehran, Iran Dispatching Management, National Iranian Gas Company, Tehran, Iran
a r t i c l e i n f o
a b s t r a c t
Article history: Received 26 May 2014 Received in revised form 13 July 2014 Accepted 14 July 2014 Available online
Pipeline networks transmit gas between the source of natural gas and customers. The high cost of transportation is a challenge that must be addressed to decrease costs. Optimizing transmission operations can decrease energy consumption in compressor stations of the network. Considering significant fuel consumption in network, the present study examined common methods of optimizing fuel consumption in gas transmission networks and then simulated a simple network model using Simone simulation software. The theoretical effects of the methods were then applied to the National Iranian Gas Network, which currently transfers gas using 65 compressor stations and 33,000 km of high pressure pipeline. The methods studied were: use of maximum and balanced capacity of pipelines; adjusting the optimum inlet pressure at city inlets, industries and power plants; use of appropriate connections to assist gas transportation flow and decrease compressor load; and selecting appropriate numbers of in-service compressors to handle the volume of gas transferred. Results indicate that substantial cost savings can be realized from the decrease in gas consumption of turbo-compressors that will also postpone overhaul time by decreasing the hours of operation. © 2014 Elsevier B.V. All rights reserved.
Keywords: Gas pipeline Optimization Network operation Simulation Fuel consumption
1. Introduction Natural gas is becoming one of the most widely used sources of energy in the world due to it's low price and environmental friendly characteristics. Based on latest BP review, Iran, Russia and Qatar hold around half of the worlds proved gas reservoirs (British Petroleum, 2014). Estimations indicate that Iran owns 16.8% of global gas reservoirs which is the largest one. Usually the location of natural gas resources and the place where the gas needed for various applications are far apart. In order to overcome the problem of long distance two common ways are widely suggested: The application of pipeline networks and Liquefied Natural Gas (LNG). As reported in Ibrahim et al. (2000), short distances gas transportation by pipelines is more economical than LNG transportation. In this work pipeline networks are chosen.
* Corresponding author. Department of Chemical Engineering, Faculty of Technology and Engineering, Tarbiat Modares University, Tehran, Iran. Fax: þ98 21 82883314. E-mail addresses: [email protected], maryam_fasihizadeh@yahoo. com (M. Fasihizadeh), [email protected] (M.V. Sefti), [email protected] (H.M. Torbati). http://dx.doi.org/10.1016/j.jngse.2014.07.018 1875-5100/© 2014 Elsevier B.V. All rights reserved.
Most gas reserves are located in the southern portions of Iran, with the exception of Khangiran field in northeastern Iran and several small fields in central and western Iran. The National Iranian Gas Company yearly transmits 170 billion m3 of natural gas from gas plants in southern Iran to consumers across the country using 33,000 km of pipeline and 65 gas stations housing 233 turbocompressor units. The main sources of consumption are:
Residences, retail businesses, and small industry Power plants Large industry Injection into oil reservoirs or gas storage Export.
Fig. 1 shows that most gas is used in residences, retail businesses, and small industry (>50%). Large industry and power plants are the next largest users. Residences, retail businesses, and small industry show different patterns of usage than the other consumers. For example, consumption increases from 110 MMSCMD in summer to 450 MMSCMD in winter in response to the need for heating. By contrast, peak consumption for power plants is in summer, when power is needed for cooling systems; consumption by power plants
320
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
nchez and Rios-Mercado, 2005; Wu et al., 2000). This gas (Borraz-Sa is actually a huge amount of gas especially for the network transmitting large volume of gas. Investigation on various pipeline network indicated that the overall operating cost of the system is highly dependent upon the operating cost of compressor stations which represents between 25% and 50% of the total company's operating budget (Rios-Mercado et al., 2006). Energy subsidies are gradually being phased out at all levels of industry in line with government policies. This will require large industry, such as the National Iranian Gas Company, to implement energy conservation programs to conserve fuel and reduce costs. Decreasing fuel and power consumption requires optimal network performance, but methods of increasing efficiency must consider network limitations to prevent disruptions in the transmission and distribution of gas. If the capacity of the network is factored into its construction and expansion, it must be based on accommodating the maximum supply required for cold weather. This means that, although the system will not run at capacity for about 8 months of the year, it must be ready to meet the critical demand of heating residences and businesses during winter. This provides an opportunity to develop an effective energy conservation program.
2.1. Constraints and guidelines for energy conservation Fig. 1. Gas distribution in 2012.
decreases in winter. There is less fluctuation for other consumers by season, since there is no dependence on temperature. Fluctuation in residential consumption is shown in Fig. 2. 2. Energy conservation programs The increasing number of natural gas consumers across the country requires construction and expansion of the gas transmission network. The complex system of high-pressure pipelines spanning the country requires compressor facilities to be installed at intervals to prevent drops in pressure. This extensive equipment uses a lot of energy. It is estimated that 3e5% of the gas transported is consumed by the compressors in order to compensate for the lost pressure of the
The main network constraints for the expansion and energy conservation project are:
To provide adequate pressure at delivery points Ability to meet maximum gas demand Maximum gas production at resource points Maintain compressor stations under peak conditions
operational
When it is confirmed that no conflict exists between the project and the limitations, research must determine project efficiency to decrease energy consumption. Two methods exist to increase the overall efficiency of the networks. The first is to install equipment for optimization (hardware solutions). The second is to provide management scenarios (software solutions).
Fig. 2. Fluctuation in residential consumption.
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
321
The software solution was investigated in the present study because it does not require installation of additional equipment. The optimization of pipeline network has been interesting for many researchers using numerical and analytical calculations based on the type of problem. These studies differ in decision setting and physical models. Optimization of gas pipeline can be done statically or dynamically. Optimization models for mid-term planning and contracting purpose do not require dynamics information so steady state model are appropriate. For real time operations, system dynamics must be captured in order to ensure feasible and implementable policies (Zavala, 2014). Tabkhi classified several types of pipeline network optimization methods (Tabkhi, 2007). The proposed plans are defined as steady state model and classified by the operational constraints listed above. Simulations are then carried out to test the feasibility of implementation. Practical examples were tested using Simone pipeline simulation software to evaluate the impact of methods on the network.
editing environment. Scenarios are then defined for each condition using the modeling assumptions. It is necessary to mention that simulation has been done to present effects of arrangement changes on fuel and power consumption of turbo compressors and also correctness of recommended methods performance. In the present study, objective function includes fuel and power consumption of turbo compressors and line pack fluctuations have not been considered due to following reasons:
3. Modeling
3.2. Standard conditions
Simulation has contributed significant achievements in analyzing the pipeline network problems (Hoeven, 2003). Pipeline simulation is used to determine the design and operating variables of the pipeline network for various configuration (Woldeyohannes and Abd Majid, 2011). Using the simulation software, the simple network is mapped in the edit section and different scenarios are defined in the run environment. The program is executed bearing in mind the following modeling assumptions for the gas transmission network:
Scenario 1: All compressors on both pipelines are operating and all interconnections are closed (Fig. 3).
Compressors are located at 140 km intervals Network inlet gas pressure (from source) is 1050 psia Minimum suction pressure is 700 psia Maximum discharge pressure is 1050 psia Pipeline roughness is 0.0007 inch (to calculate pressure) Gas velocity in the pipeline is <13 m/s Minimum suction pressure of gas compressors and maximum gas velocity in the pipeline reflect actual conditions An average efficiency has been used for simulating turbo compressors After running the program, modeling results are examined as tables and graphs. The most important results are values for volume of gas transferred, pressure at the end of the pipeline, compressor power and fuel consumption, compressor suction pressure and gas velocity. 3.1. Simulation stages 3.1.1. Using Simone simulation software for simulation Case study simulations: Two 700 km parallel connected pipelines with compressors installed at 140 km intervals are mapped in the
Line pack volume is related to fuel gas consumption. Line pack volume of Iran network has no considerable fluctuation in warm and mild seasons, consumption in residence and retail business is nearly constant, and other sections such as power plants and large industries don't face with unpredictable changes. As network and line pack volume are very large, small changes in gas consumption managed by pack line volume.
3.3. Optimizing fuel consumption The methods proposed to optimize fuel consumption in the gas transmission network are: Use the potential and maximum capacity of the pipelines Adjust the optimum inlet pressure at the city inlets and for industry and power plants Use the appropriate interconnections to assist gas flow and decrease compressor loads Select the appropriate number of in-service compressors
3.3.1. Maximum and balanced capacity of pipelines Optimal use of potential pipeline capacity for gas transmission increases the useful lifetime of equipment by decreasing the load and the in-service time, which decreases fuel and power consumption at the gas compressor stations. The effect of symmetric distribution of compressors on network performance is checked for 3 scenarios: Scenario 1: All compressors on one pipeline work and those on the other pipeline are bypassed. Power consumption is 184.1 MW and fuel consumption is 1.55 MMSCMD to transfer 127 MMSCMD of natural gas (Fig. 4). Scenario 2: Assume a symmetric arrangement for both pipelines using 4 compressors. Power consumption is 111 MW, fuel consumption is 0.94 MMSCMD, and pressure at the end of the network is 700 psia (Fig. 5).
Fig. 3. Simulation of standard conditions.
322
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fig. 4. Scenario no. 1-1.
Fig. 5. Scenario no. 2-1.
Scenario 3: Assume a symmetric arrangement for the two pipelines. Power consumption is 96 MW, fuel consumption is 0.81 MMSCMD, and pressure at the end of the network is 700 psia (Fig. 6). The results of these evaluations indicate that a considerable decrease in fuel and power consumption occurs when the compressors are symmetrically distributed between the two pipelines (scenarios 2 and 3) over the conditions in scenario 1. It is also clear that different options may exist to arrange the symmetry. Scenario 3 was found to consume the least fuel and power of the three choices. 3.3.2. Providing and adjusting optimum inlet pressure for cities, industry and power plants The main goal in such a complicated network is to be able to transfer gas under adequate pressure to the end of the network. Consumers along the line (cities, major industry and power plants) require a number of branches from the high pressure pipelines. When entering the distribution system from the pipeline, gas
pressure must decrease at the city gate stations. It is clear that large differential pressure between the pipeline and the desirable user pressure is a source of energy loss. At some stations, energy loss can be decreased using throttling valves instead of expansion turbines; this is not discussed here because such a project required installation of additional equipment. Fixed positions are assumed for all consumers and gas stations, so that network energy consumption can be optimized by changing the arrangement of the compressors so that pressure decreases before it reaches the delivery point. To test this assumption, two hypothetical consumers are added to the network and two scenarios using the same number of compressors with different configurations were tested. To achieve comparable results, symmetry was assumed and the pressure remained unchanged at the end of the network at 700 psia. Scenario 1: Higher pressure (917 psia) is maintained at the branch delivery points placed beyond the in-service compressors. To transfer 140 MMSCMD of gas to the end of the network requires 124 MW of power and 1.05 MMSCMD of fuel (Fig. 7).
Fig. 6. Scenario no. 3-1.
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
323
Fig. 7. Scenario no. 1-2.
Scenario 2: Lower pressure is achieved at one branch by changing the arrangement of the compressors at delivery points located beyond the by-passed compressors. To transfer 140 MMSCMD of gas to end of the network, 122 MW of power and 1.03 MMSCMD of fuel is needed (Fig. 8). The results showed that a proper compressor arrangement produces lower pressure at the branches and decreases pressure loss. Although, power and fuel consumption decrease slightly, the overall effect is considerable for an extended network with many consumers, such as the Iranian Gas Trunk.
3.3.3. Use of appropriate interconnections to decrease compressors load Many interconnections exist along the pipelines as branching to users occurs. Effective use of the interconnections can improve network efficiency. Improper use of interconnections can decrease efficiency and increase energy loss; thus, selecting appropriate arrangements for interconnections (closing and opening) under different operational conditions, especially for sudden changes in pressure, is a major function of network optimization management. The following are assumed: Network inlet flow is 150 MMSCMD The size of deliveries is 23 and 25 MMSCMD Network outlet flow is 101 MMSCMD
Three scenarios are designed to test the effects of interconnections on the network for two consumers. Scenario 1: All interconnections are closed and the gas flows in two independent parallel pipelines. Study of possible compressor arrangements requiring minimum power indicates that 3 compressors should be by-passed (nos. 3, 4, 8). Using 113 MW of power and 0.95 MMSCMD of fuel, 150 MMSCMD of gas can be transmitted to the network end point (Fig. 9). Scenario 2: Open two interconnections between the inlets of compressors 3 and 7 (valve no. 5), which decreases power consumption about 2 MW. Fuel consumption is 0.93 MMSCMD. Scenario 3: The interconnections between inlets of compressors 2 and 6 (valve no. 3) and 3 and 7 (valve no. 5) are opened. Power consumption decreases about 10 MWe101 MW. Fuel consumption decreases to 0.85 MMSCMD and pressure at the network end point remains unchanged (Table 1). Evaluation of results indicates that scenario 3 has a positive effect on decreasing energy consumption (power and fuel). Opening some interconnections, such as the ones between inlets of compressors 1 and 5, has no effect on the network. Appropriate opening of interconnections can decrease power and fuel consumption, but improper openings can increase power consumption and cause flow to recycle back into the pipeline. For example, when parallel compressors exist and one is open while the other is bypassed, the discharge interconnection should not be opened because recycling will occur.
Fig. 8. Scenario no. 2-2.
324
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fig. 9. Scenario no. 1-3.
Table 1 Different scenario for using of appropriate interconnections.
Scenario no. 1-3 Scenario no. 2-3 Scenario no. 3-3
Valve 1
Valve 2
Valve 3
Valve 4
Valve 5
Valve 6
P1
P2
Power
Fuel
Closed Closed Closed
Closed Closed Closed
Closed Closed Open
Closed Closed Closed
Closed Open Open
Closed Closed Closed
900 832 880
885 871 870
113 111 101
0.95 0.93 0.85
3.3.4. Selecting proper number of compressors for network All compressors should be in service to maintain the network in operational condition and maintain flow. In the simulated network, the maximum allowable flow is 179 MMSCMD, fuel consumption is 3.3 MMSCMD, and power consumption is 388.1 MW. It is not necessary for all compressors to operate continually when the transfer flow is low. Deciding which compressors should be operational depends on the amount that power and fuel consumption can be decreased while respecting network constraints. To test this hypothesis in the simulated network, inlet gas was decreased from its maximum flow of 179 MMSCMDe150 MMSCMD. At this rate, it is not necessary for all compressors to be in service; by keeping operational conditions constant, new scenarios can be defined. Scenario 1: 7 compressors remain in service to transfer gas using 188 MW of power and 1.6 MMSCMD of fuel to obtain 700 psia of pressure at the end of the network (Fig. 10). Scenario 2: 6 compressors remain in service using 214 MW of power and 1.8 MMSCMD of fuel with the pressure remaining unchanged at the end of the network (Fig. 11). The results indicate that decreasing the transfer flow in Scenario 1 eliminates the need for all compressors to be in service; however,
reducing the number of compressors in service did not produce optimal conditions. Scenario 2 was defined as a means of comparison. A comparison of the two scenarios shows that using fewer compressors and maintaining desirable pressure at the end of network in scenario 2 increased the load on the other in-service compressors and increased the total power consumed, which are not optimal conditions. This suggests that choosing the appropriate number of in-service compressors should consider all operational aspects. 4. Optimizing Iranian gas transmission network After evaluating simulation outputs and verifying accuracy of these methods using simulation algorithms, different methods applied for Iranian gas transmission network for two years, separately and in combination. Before applying each one of recommended methods, facing any problem originated from network managing decisions prevented by getting exact information about production and consumption plans in the near future. Then appropriate arrangement selected using mentioned rules. After that Simone software utilized for evaluating it and at last, recommendations implied in network by dispatching control center after informing different network disciplines.
Fig. 10. Scenario no. 1-4.
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
325
Fig. 11. Scenario no. 2-4.
It is necessary to mention that according to last published statistics, Iran Network transferred about 600 MMSCMD in 2013 from gas plants to different consumers, while line pack been about 1,000,000,000 MMSCMD, so instantaneous fluctuations in production or consumption neutralized and do not have considerable effect on the number of gas compressor stations and network arrangement. In addition, as arrangement changes are done consciously and considering overhaul plans and last for two or three weeks, there is no need to shut down and start up compressors continuously and consequently, no more cost should be paid by network. Results compared for fuel consumption and average amount of gas transmitted. Real data used for regional gas transmission. It is daily mean value for different months from 2009 to 2011 were: Power consumption and operational hours for turbo compressors at gas stations Fuel consumption at stations Volume of gas transmitted and consumed and the balance of production versus consumption for each region Length of the trunk line
The average amount of gas transmitted (ðQ Þnetwork ) is multiplied by the average length of the network (ðLÞnetwork ) and ðQ Þnetwork ðLÞnetwork is compared for different months. This was carried out for three years (2009 through 2011) (Fig. 12). Fig. 12 shows that the amount of gas transmitted over 1000 km of pipeline increased and that network capacity in 2011 was greater than for 2009. Fig. 13 shows that, despite the 8.4% increase in network transmission potential in 2010 over 2009, fuel consumption per 1 MMSCMD of transferred gas and the hours of operation decreased. It is clear that the recommended guidelines were useful and that results further improved in 2011. The amount of gas transmitted in 2011 increased by about 12.4% although the fuel consumption per 1 MMSCMD of gas transferred and hours of operation decreased (Table 2). Observed differences in results can be interpreted as following: Iranian gas network changed from production and consumption point of view in 2011 in comparison with 2010 as some new turbo compressors and pipelines added to network, however, new facilities are negligible when compared to whole network. In addition, generally, number and duration of overhauls for pipelines and gas compressor stations are different in various years.
Fig. 12. ðQ Þnetwork ðLÞnetwork for different months.
326
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fig. 13. Amount of gas transmitted, fuel consumed, and hours of operation of compressors in 2009e2011.
Table 2 Comparison real data between 2009 and 2011.
Average volume of transferred gas (Q) MMSCMD Length of transmission (km) Fuel consumption per transferred 1 MMSCMD gas at 1000 km Operational hour of gas station (h)
2009
2010
2011
394 764 1.39%
427 803 1.21%
443 806 1.12%
1484
1416
1301
5. Conclusion The substantial expansion of the gas transmission network in Iran will increase the amount of energy consumed to transmit the natural gas through the ever-expanding network. The National Iranian Gas Company requires programs to decrease energy consumption and increase efficiency of transmission. Most plans require installation of new equipment to improve efficiency, which requires an initial outlay of capital. The amount of decrease in energy consumption for this initial investment over a specified period of time may not warrant the purchase and installation of expensive equipment. This slows implementation and delays the commencement of energy-saving measures. The present study tested 4 models to increase the capacity of the network and decrease transmission costs that improve the gas transmission network without an initial capital outlay. The accuracy of these methods was verified using simulation algorithms for different cases, optimizing Iranian gas transmission network and calculating the results of implementation of the models for 2010 and 2011 for over 33,000 km of high pressure pipeline and 69 stations. The results are summarized below. (1) Despite the increase in network capacity, energy consumption decreased significantly in 2010 and 2011 in comparison with 2009. Fuel consumption per specified volume of transported gas decreased to 1.21% in 2010 and 1.12% in 2011 compared to the 1.39% reported for 2009. This difference was greater during warmer months when factoring in network potential.
(2) Hours of operation of the compressors decreased noticeably for the recommended models. The increased overhaul intervals decreased operational costs and prolonged the lifetimes of the equipment. (3) The estimated cost savings for fuel consumption was $50,573,000 in 2010 and $51,576,000 for 2011 when calculated at $300/1000 m3. (4) Observed differences in results of applying proposed methods can be interpreted as following: a. Changing the number and duration of overhauls for pipelines and gas compressor stations b. Installing new facilities in various years. (5) These methods modify network performance considerably, but for approaching precise scenarios and complete optimization of network, optimization algorithms (classic mathematical and meta heuristic methods) should be studied thoroughly and compared with models recommended in the present research to identify pros and cons. (6) The Iranian Gas Network is sophisticated and specialized; optimization methods should be tested on a small section of the network under the same specifications applied to the entire network to verify its application for all sections. This case is under study and primary test, which its result will be published in future papers for one of the important areas of Iranian gas network. Acknowledgment The authors would like to thank N.I.G.C (National Iranian Gas Company) (Contract no: 192002) for supporting this work and The Dispatching Management of N.I.G.C for providing useful data. Nomenclature Q average of transferred gas volume P pressure of supply or delivery MMSCMD million standard cubic meter per day LNG liquefied natural gas
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fuel m/s MW
gas consumption in compressor station meter per second (velocity unit) mega watt (power unit)
References nchez, C., Rios-Mercado, R.Z., 2005. A hybrid metha-heuristic approach for Borraz-Sa natural gas pipeline network optimization. In: Hybrid Metha Heuristics. Springer, pp. 54e65. British Petroleum (BP), June 2014. Statistical Review of World Energy. David Wood & Associates. http://www.bp.com/statisticalreview. Hoeven, T.V.D., 2003. Constrained network simulation. In: The 35th Annual Meeting of Pipeline Simulation Interest Group (PSIG). Bern, Switzerland.
327
Ibrahim, H., Bin Yaman, A.S., Thanh Lien, T., Budi Santoso, I., Laothumthut, P., 2000. Natural Gas Pipeline Development in South East Asia. Asia Pacific Energy Research Center, Tokyo. Rios-Mercado, R.Z., Kim, S., Boyd, E.A., 2006. Efficient operation of natural gas transmission systems: a network based heuristic for cyclic structures. Comput. Operat. Res. 33, 2323e2351. Tabkhi, F., 2007. Optimization of Gas Transmission Networks (Ph.D. thesis). INPT. Woldeyohannes, A.D., Abd Majid, M.A., 2011. Simulation model for natural gas transmission pipeline network system. Simul. Model. Pract. Theory 19, 196e212. Wu, S., Rios-Mercado, R.Z., Boyd, E.A., Scott, L.R., 2000. Model relaxations for the fuel cost minimization of steady-state gas pipeline networks. Math. Comput. Model. 31, 197e220. Zavala, V.M., 2014. Stochastic optimal control model for natural gas networks. Comput. Chem. Eng. 64, 103e113.
Contents lists available at ScienceDirect
Journal of Natural Gas Science and Engineering journal homepage: www.elsevier.com/locate/jngse
Improving gas transmission networks operation using simulation algorithms: Case study of the National Iranian Gas Network Maryam Fasihizadeh a, Mohsen V. Sefti a, *, Hassan M. Torbati b a b
Department of Chemical Engineering, Tarbiat Modares University, Tehran, Iran Dispatching Management, National Iranian Gas Company, Tehran, Iran
a r t i c l e i n f o
a b s t r a c t
Article history: Received 26 May 2014 Received in revised form 13 July 2014 Accepted 14 July 2014 Available online
Pipeline networks transmit gas between the source of natural gas and customers. The high cost of transportation is a challenge that must be addressed to decrease costs. Optimizing transmission operations can decrease energy consumption in compressor stations of the network. Considering significant fuel consumption in network, the present study examined common methods of optimizing fuel consumption in gas transmission networks and then simulated a simple network model using Simone simulation software. The theoretical effects of the methods were then applied to the National Iranian Gas Network, which currently transfers gas using 65 compressor stations and 33,000 km of high pressure pipeline. The methods studied were: use of maximum and balanced capacity of pipelines; adjusting the optimum inlet pressure at city inlets, industries and power plants; use of appropriate connections to assist gas transportation flow and decrease compressor load; and selecting appropriate numbers of in-service compressors to handle the volume of gas transferred. Results indicate that substantial cost savings can be realized from the decrease in gas consumption of turbo-compressors that will also postpone overhaul time by decreasing the hours of operation. © 2014 Elsevier B.V. All rights reserved.
Keywords: Gas pipeline Optimization Network operation Simulation Fuel consumption
1. Introduction Natural gas is becoming one of the most widely used sources of energy in the world due to it's low price and environmental friendly characteristics. Based on latest BP review, Iran, Russia and Qatar hold around half of the worlds proved gas reservoirs (British Petroleum, 2014). Estimations indicate that Iran owns 16.8% of global gas reservoirs which is the largest one. Usually the location of natural gas resources and the place where the gas needed for various applications are far apart. In order to overcome the problem of long distance two common ways are widely suggested: The application of pipeline networks and Liquefied Natural Gas (LNG). As reported in Ibrahim et al. (2000), short distances gas transportation by pipelines is more economical than LNG transportation. In this work pipeline networks are chosen.
* Corresponding author. Department of Chemical Engineering, Faculty of Technology and Engineering, Tarbiat Modares University, Tehran, Iran. Fax: þ98 21 82883314. E-mail addresses: [email protected], maryam_fasihizadeh@yahoo. com (M. Fasihizadeh), [email protected] (M.V. Sefti), [email protected] (H.M. Torbati). http://dx.doi.org/10.1016/j.jngse.2014.07.018 1875-5100/© 2014 Elsevier B.V. All rights reserved.
Most gas reserves are located in the southern portions of Iran, with the exception of Khangiran field in northeastern Iran and several small fields in central and western Iran. The National Iranian Gas Company yearly transmits 170 billion m3 of natural gas from gas plants in southern Iran to consumers across the country using 33,000 km of pipeline and 65 gas stations housing 233 turbocompressor units. The main sources of consumption are:
Residences, retail businesses, and small industry Power plants Large industry Injection into oil reservoirs or gas storage Export.
Fig. 1 shows that most gas is used in residences, retail businesses, and small industry (>50%). Large industry and power plants are the next largest users. Residences, retail businesses, and small industry show different patterns of usage than the other consumers. For example, consumption increases from 110 MMSCMD in summer to 450 MMSCMD in winter in response to the need for heating. By contrast, peak consumption for power plants is in summer, when power is needed for cooling systems; consumption by power plants
320
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
nchez and Rios-Mercado, 2005; Wu et al., 2000). This gas (Borraz-Sa is actually a huge amount of gas especially for the network transmitting large volume of gas. Investigation on various pipeline network indicated that the overall operating cost of the system is highly dependent upon the operating cost of compressor stations which represents between 25% and 50% of the total company's operating budget (Rios-Mercado et al., 2006). Energy subsidies are gradually being phased out at all levels of industry in line with government policies. This will require large industry, such as the National Iranian Gas Company, to implement energy conservation programs to conserve fuel and reduce costs. Decreasing fuel and power consumption requires optimal network performance, but methods of increasing efficiency must consider network limitations to prevent disruptions in the transmission and distribution of gas. If the capacity of the network is factored into its construction and expansion, it must be based on accommodating the maximum supply required for cold weather. This means that, although the system will not run at capacity for about 8 months of the year, it must be ready to meet the critical demand of heating residences and businesses during winter. This provides an opportunity to develop an effective energy conservation program.
2.1. Constraints and guidelines for energy conservation Fig. 1. Gas distribution in 2012.
decreases in winter. There is less fluctuation for other consumers by season, since there is no dependence on temperature. Fluctuation in residential consumption is shown in Fig. 2. 2. Energy conservation programs The increasing number of natural gas consumers across the country requires construction and expansion of the gas transmission network. The complex system of high-pressure pipelines spanning the country requires compressor facilities to be installed at intervals to prevent drops in pressure. This extensive equipment uses a lot of energy. It is estimated that 3e5% of the gas transported is consumed by the compressors in order to compensate for the lost pressure of the
The main network constraints for the expansion and energy conservation project are:
To provide adequate pressure at delivery points Ability to meet maximum gas demand Maximum gas production at resource points Maintain compressor stations under peak conditions
operational
When it is confirmed that no conflict exists between the project and the limitations, research must determine project efficiency to decrease energy consumption. Two methods exist to increase the overall efficiency of the networks. The first is to install equipment for optimization (hardware solutions). The second is to provide management scenarios (software solutions).
Fig. 2. Fluctuation in residential consumption.
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
321
The software solution was investigated in the present study because it does not require installation of additional equipment. The optimization of pipeline network has been interesting for many researchers using numerical and analytical calculations based on the type of problem. These studies differ in decision setting and physical models. Optimization of gas pipeline can be done statically or dynamically. Optimization models for mid-term planning and contracting purpose do not require dynamics information so steady state model are appropriate. For real time operations, system dynamics must be captured in order to ensure feasible and implementable policies (Zavala, 2014). Tabkhi classified several types of pipeline network optimization methods (Tabkhi, 2007). The proposed plans are defined as steady state model and classified by the operational constraints listed above. Simulations are then carried out to test the feasibility of implementation. Practical examples were tested using Simone pipeline simulation software to evaluate the impact of methods on the network.
editing environment. Scenarios are then defined for each condition using the modeling assumptions. It is necessary to mention that simulation has been done to present effects of arrangement changes on fuel and power consumption of turbo compressors and also correctness of recommended methods performance. In the present study, objective function includes fuel and power consumption of turbo compressors and line pack fluctuations have not been considered due to following reasons:
3. Modeling
3.2. Standard conditions
Simulation has contributed significant achievements in analyzing the pipeline network problems (Hoeven, 2003). Pipeline simulation is used to determine the design and operating variables of the pipeline network for various configuration (Woldeyohannes and Abd Majid, 2011). Using the simulation software, the simple network is mapped in the edit section and different scenarios are defined in the run environment. The program is executed bearing in mind the following modeling assumptions for the gas transmission network:
Scenario 1: All compressors on both pipelines are operating and all interconnections are closed (Fig. 3).
Compressors are located at 140 km intervals Network inlet gas pressure (from source) is 1050 psia Minimum suction pressure is 700 psia Maximum discharge pressure is 1050 psia Pipeline roughness is 0.0007 inch (to calculate pressure) Gas velocity in the pipeline is <13 m/s Minimum suction pressure of gas compressors and maximum gas velocity in the pipeline reflect actual conditions An average efficiency has been used for simulating turbo compressors After running the program, modeling results are examined as tables and graphs. The most important results are values for volume of gas transferred, pressure at the end of the pipeline, compressor power and fuel consumption, compressor suction pressure and gas velocity. 3.1. Simulation stages 3.1.1. Using Simone simulation software for simulation Case study simulations: Two 700 km parallel connected pipelines with compressors installed at 140 km intervals are mapped in the
Line pack volume is related to fuel gas consumption. Line pack volume of Iran network has no considerable fluctuation in warm and mild seasons, consumption in residence and retail business is nearly constant, and other sections such as power plants and large industries don't face with unpredictable changes. As network and line pack volume are very large, small changes in gas consumption managed by pack line volume.
3.3. Optimizing fuel consumption The methods proposed to optimize fuel consumption in the gas transmission network are: Use the potential and maximum capacity of the pipelines Adjust the optimum inlet pressure at the city inlets and for industry and power plants Use the appropriate interconnections to assist gas flow and decrease compressor loads Select the appropriate number of in-service compressors
3.3.1. Maximum and balanced capacity of pipelines Optimal use of potential pipeline capacity for gas transmission increases the useful lifetime of equipment by decreasing the load and the in-service time, which decreases fuel and power consumption at the gas compressor stations. The effect of symmetric distribution of compressors on network performance is checked for 3 scenarios: Scenario 1: All compressors on one pipeline work and those on the other pipeline are bypassed. Power consumption is 184.1 MW and fuel consumption is 1.55 MMSCMD to transfer 127 MMSCMD of natural gas (Fig. 4). Scenario 2: Assume a symmetric arrangement for both pipelines using 4 compressors. Power consumption is 111 MW, fuel consumption is 0.94 MMSCMD, and pressure at the end of the network is 700 psia (Fig. 5).
Fig. 3. Simulation of standard conditions.
322
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fig. 4. Scenario no. 1-1.
Fig. 5. Scenario no. 2-1.
Scenario 3: Assume a symmetric arrangement for the two pipelines. Power consumption is 96 MW, fuel consumption is 0.81 MMSCMD, and pressure at the end of the network is 700 psia (Fig. 6). The results of these evaluations indicate that a considerable decrease in fuel and power consumption occurs when the compressors are symmetrically distributed between the two pipelines (scenarios 2 and 3) over the conditions in scenario 1. It is also clear that different options may exist to arrange the symmetry. Scenario 3 was found to consume the least fuel and power of the three choices. 3.3.2. Providing and adjusting optimum inlet pressure for cities, industry and power plants The main goal in such a complicated network is to be able to transfer gas under adequate pressure to the end of the network. Consumers along the line (cities, major industry and power plants) require a number of branches from the high pressure pipelines. When entering the distribution system from the pipeline, gas
pressure must decrease at the city gate stations. It is clear that large differential pressure between the pipeline and the desirable user pressure is a source of energy loss. At some stations, energy loss can be decreased using throttling valves instead of expansion turbines; this is not discussed here because such a project required installation of additional equipment. Fixed positions are assumed for all consumers and gas stations, so that network energy consumption can be optimized by changing the arrangement of the compressors so that pressure decreases before it reaches the delivery point. To test this assumption, two hypothetical consumers are added to the network and two scenarios using the same number of compressors with different configurations were tested. To achieve comparable results, symmetry was assumed and the pressure remained unchanged at the end of the network at 700 psia. Scenario 1: Higher pressure (917 psia) is maintained at the branch delivery points placed beyond the in-service compressors. To transfer 140 MMSCMD of gas to the end of the network requires 124 MW of power and 1.05 MMSCMD of fuel (Fig. 7).
Fig. 6. Scenario no. 3-1.
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
323
Fig. 7. Scenario no. 1-2.
Scenario 2: Lower pressure is achieved at one branch by changing the arrangement of the compressors at delivery points located beyond the by-passed compressors. To transfer 140 MMSCMD of gas to end of the network, 122 MW of power and 1.03 MMSCMD of fuel is needed (Fig. 8). The results showed that a proper compressor arrangement produces lower pressure at the branches and decreases pressure loss. Although, power and fuel consumption decrease slightly, the overall effect is considerable for an extended network with many consumers, such as the Iranian Gas Trunk.
3.3.3. Use of appropriate interconnections to decrease compressors load Many interconnections exist along the pipelines as branching to users occurs. Effective use of the interconnections can improve network efficiency. Improper use of interconnections can decrease efficiency and increase energy loss; thus, selecting appropriate arrangements for interconnections (closing and opening) under different operational conditions, especially for sudden changes in pressure, is a major function of network optimization management. The following are assumed: Network inlet flow is 150 MMSCMD The size of deliveries is 23 and 25 MMSCMD Network outlet flow is 101 MMSCMD
Three scenarios are designed to test the effects of interconnections on the network for two consumers. Scenario 1: All interconnections are closed and the gas flows in two independent parallel pipelines. Study of possible compressor arrangements requiring minimum power indicates that 3 compressors should be by-passed (nos. 3, 4, 8). Using 113 MW of power and 0.95 MMSCMD of fuel, 150 MMSCMD of gas can be transmitted to the network end point (Fig. 9). Scenario 2: Open two interconnections between the inlets of compressors 3 and 7 (valve no. 5), which decreases power consumption about 2 MW. Fuel consumption is 0.93 MMSCMD. Scenario 3: The interconnections between inlets of compressors 2 and 6 (valve no. 3) and 3 and 7 (valve no. 5) are opened. Power consumption decreases about 10 MWe101 MW. Fuel consumption decreases to 0.85 MMSCMD and pressure at the network end point remains unchanged (Table 1). Evaluation of results indicates that scenario 3 has a positive effect on decreasing energy consumption (power and fuel). Opening some interconnections, such as the ones between inlets of compressors 1 and 5, has no effect on the network. Appropriate opening of interconnections can decrease power and fuel consumption, but improper openings can increase power consumption and cause flow to recycle back into the pipeline. For example, when parallel compressors exist and one is open while the other is bypassed, the discharge interconnection should not be opened because recycling will occur.
Fig. 8. Scenario no. 2-2.
324
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fig. 9. Scenario no. 1-3.
Table 1 Different scenario for using of appropriate interconnections.
Scenario no. 1-3 Scenario no. 2-3 Scenario no. 3-3
Valve 1
Valve 2
Valve 3
Valve 4
Valve 5
Valve 6
P1
P2
Power
Fuel
Closed Closed Closed
Closed Closed Closed
Closed Closed Open
Closed Closed Closed
Closed Open Open
Closed Closed Closed
900 832 880
885 871 870
113 111 101
0.95 0.93 0.85
3.3.4. Selecting proper number of compressors for network All compressors should be in service to maintain the network in operational condition and maintain flow. In the simulated network, the maximum allowable flow is 179 MMSCMD, fuel consumption is 3.3 MMSCMD, and power consumption is 388.1 MW. It is not necessary for all compressors to operate continually when the transfer flow is low. Deciding which compressors should be operational depends on the amount that power and fuel consumption can be decreased while respecting network constraints. To test this hypothesis in the simulated network, inlet gas was decreased from its maximum flow of 179 MMSCMDe150 MMSCMD. At this rate, it is not necessary for all compressors to be in service; by keeping operational conditions constant, new scenarios can be defined. Scenario 1: 7 compressors remain in service to transfer gas using 188 MW of power and 1.6 MMSCMD of fuel to obtain 700 psia of pressure at the end of the network (Fig. 10). Scenario 2: 6 compressors remain in service using 214 MW of power and 1.8 MMSCMD of fuel with the pressure remaining unchanged at the end of the network (Fig. 11). The results indicate that decreasing the transfer flow in Scenario 1 eliminates the need for all compressors to be in service; however,
reducing the number of compressors in service did not produce optimal conditions. Scenario 2 was defined as a means of comparison. A comparison of the two scenarios shows that using fewer compressors and maintaining desirable pressure at the end of network in scenario 2 increased the load on the other in-service compressors and increased the total power consumed, which are not optimal conditions. This suggests that choosing the appropriate number of in-service compressors should consider all operational aspects. 4. Optimizing Iranian gas transmission network After evaluating simulation outputs and verifying accuracy of these methods using simulation algorithms, different methods applied for Iranian gas transmission network for two years, separately and in combination. Before applying each one of recommended methods, facing any problem originated from network managing decisions prevented by getting exact information about production and consumption plans in the near future. Then appropriate arrangement selected using mentioned rules. After that Simone software utilized for evaluating it and at last, recommendations implied in network by dispatching control center after informing different network disciplines.
Fig. 10. Scenario no. 1-4.
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
325
Fig. 11. Scenario no. 2-4.
It is necessary to mention that according to last published statistics, Iran Network transferred about 600 MMSCMD in 2013 from gas plants to different consumers, while line pack been about 1,000,000,000 MMSCMD, so instantaneous fluctuations in production or consumption neutralized and do not have considerable effect on the number of gas compressor stations and network arrangement. In addition, as arrangement changes are done consciously and considering overhaul plans and last for two or three weeks, there is no need to shut down and start up compressors continuously and consequently, no more cost should be paid by network. Results compared for fuel consumption and average amount of gas transmitted. Real data used for regional gas transmission. It is daily mean value for different months from 2009 to 2011 were: Power consumption and operational hours for turbo compressors at gas stations Fuel consumption at stations Volume of gas transmitted and consumed and the balance of production versus consumption for each region Length of the trunk line
The average amount of gas transmitted (ðQ Þnetwork ) is multiplied by the average length of the network (ðLÞnetwork ) and ðQ Þnetwork ðLÞnetwork is compared for different months. This was carried out for three years (2009 through 2011) (Fig. 12). Fig. 12 shows that the amount of gas transmitted over 1000 km of pipeline increased and that network capacity in 2011 was greater than for 2009. Fig. 13 shows that, despite the 8.4% increase in network transmission potential in 2010 over 2009, fuel consumption per 1 MMSCMD of transferred gas and the hours of operation decreased. It is clear that the recommended guidelines were useful and that results further improved in 2011. The amount of gas transmitted in 2011 increased by about 12.4% although the fuel consumption per 1 MMSCMD of gas transferred and hours of operation decreased (Table 2). Observed differences in results can be interpreted as following: Iranian gas network changed from production and consumption point of view in 2011 in comparison with 2010 as some new turbo compressors and pipelines added to network, however, new facilities are negligible when compared to whole network. In addition, generally, number and duration of overhauls for pipelines and gas compressor stations are different in various years.
Fig. 12. ðQ Þnetwork ðLÞnetwork for different months.
326
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fig. 13. Amount of gas transmitted, fuel consumed, and hours of operation of compressors in 2009e2011.
Table 2 Comparison real data between 2009 and 2011.
Average volume of transferred gas (Q) MMSCMD Length of transmission (km) Fuel consumption per transferred 1 MMSCMD gas at 1000 km Operational hour of gas station (h)
2009
2010
2011
394 764 1.39%
427 803 1.21%
443 806 1.12%
1484
1416
1301
5. Conclusion The substantial expansion of the gas transmission network in Iran will increase the amount of energy consumed to transmit the natural gas through the ever-expanding network. The National Iranian Gas Company requires programs to decrease energy consumption and increase efficiency of transmission. Most plans require installation of new equipment to improve efficiency, which requires an initial outlay of capital. The amount of decrease in energy consumption for this initial investment over a specified period of time may not warrant the purchase and installation of expensive equipment. This slows implementation and delays the commencement of energy-saving measures. The present study tested 4 models to increase the capacity of the network and decrease transmission costs that improve the gas transmission network without an initial capital outlay. The accuracy of these methods was verified using simulation algorithms for different cases, optimizing Iranian gas transmission network and calculating the results of implementation of the models for 2010 and 2011 for over 33,000 km of high pressure pipeline and 69 stations. The results are summarized below. (1) Despite the increase in network capacity, energy consumption decreased significantly in 2010 and 2011 in comparison with 2009. Fuel consumption per specified volume of transported gas decreased to 1.21% in 2010 and 1.12% in 2011 compared to the 1.39% reported for 2009. This difference was greater during warmer months when factoring in network potential.
(2) Hours of operation of the compressors decreased noticeably for the recommended models. The increased overhaul intervals decreased operational costs and prolonged the lifetimes of the equipment. (3) The estimated cost savings for fuel consumption was $50,573,000 in 2010 and $51,576,000 for 2011 when calculated at $300/1000 m3. (4) Observed differences in results of applying proposed methods can be interpreted as following: a. Changing the number and duration of overhauls for pipelines and gas compressor stations b. Installing new facilities in various years. (5) These methods modify network performance considerably, but for approaching precise scenarios and complete optimization of network, optimization algorithms (classic mathematical and meta heuristic methods) should be studied thoroughly and compared with models recommended in the present research to identify pros and cons. (6) The Iranian Gas Network is sophisticated and specialized; optimization methods should be tested on a small section of the network under the same specifications applied to the entire network to verify its application for all sections. This case is under study and primary test, which its result will be published in future papers for one of the important areas of Iranian gas network. Acknowledgment The authors would like to thank N.I.G.C (National Iranian Gas Company) (Contract no: 192002) for supporting this work and The Dispatching Management of N.I.G.C for providing useful data. Nomenclature Q average of transferred gas volume P pressure of supply or delivery MMSCMD million standard cubic meter per day LNG liquefied natural gas
M. Fasihizadeh et al. / Journal of Natural Gas Science and Engineering 20 (2014) 319e327
Fuel m/s MW
gas consumption in compressor station meter per second (velocity unit) mega watt (power unit)
References nchez, C., Rios-Mercado, R.Z., 2005. A hybrid metha-heuristic approach for Borraz-Sa natural gas pipeline network optimization. In: Hybrid Metha Heuristics. Springer, pp. 54e65. British Petroleum (BP), June 2014. Statistical Review of World Energy. David Wood & Associates. http://www.bp.com/statisticalreview. Hoeven, T.V.D., 2003. Constrained network simulation. In: The 35th Annual Meeting of Pipeline Simulation Interest Group (PSIG). Bern, Switzerland.
327
Ibrahim, H., Bin Yaman, A.S., Thanh Lien, T., Budi Santoso, I., Laothumthut, P., 2000. Natural Gas Pipeline Development in South East Asia. Asia Pacific Energy Research Center, Tokyo. Rios-Mercado, R.Z., Kim, S., Boyd, E.A., 2006. Efficient operation of natural gas transmission systems: a network based heuristic for cyclic structures. Comput. Operat. Res. 33, 2323e2351. Tabkhi, F., 2007. Optimization of Gas Transmission Networks (Ph.D. thesis). INPT. Woldeyohannes, A.D., Abd Majid, M.A., 2011. Simulation model for natural gas transmission pipeline network system. Simul. Model. Pract. Theory 19, 196e212. Wu, S., Rios-Mercado, R.Z., Boyd, E.A., Scott, L.R., 2000. Model relaxations for the fuel cost minimization of steady-state gas pipeline networks. Math. Comput. Model. 31, 197e220. Zavala, V.M., 2014. Stochastic optimal control model for natural gas networks. Comput. Chem. Eng. 64, 103e113.