A part of the pipeline which is approximately 250 miles in length transports gas to a number of states of the northeastern U. S. six compressor stations along this particular segment are present there. Most of the gas enters the pipeline at the first station with the balance entering at the fifth station. Gas is delivered at numerous points along the pipeline and the remainder flows out into the next part of the pipeline. Either the suction pressure, the flow, or the discharge pressure is controlled locally at each of these stations. Fig. shows a schematic of the pipeline segment. The first station consists of six reciprocating compressors.
Two of these compressors are equipped with air/fuel ratio control and digital speed control. This station is usually on flow control to meet contractual obligations. Three gas-turbine-driven centrifugal compressors are present at the second station and suction pressure is controlled at this station. Suction pressure is controlled by maintaining turbine speed. Three gas-driven reciprocating compressors are had by the third station. This station is normally on suction control and individual engine controls are very similar to the first station. The fourth station also employs three gas-turbine-driven centrifugal compressors.
Control at this station is very similar to the second station. The fifth station consists of six reciprocating compressors and much of the delivery along the line occurs at this station. The sixth station employs two turbine-driven centrifugal compressors and normally is run for peak deliveries only. It normally is set on suction control at the time of operation. Currently an online supervisory control and data acquisition (SCADA) system is used to control the whole system centrally by using the dispatchers. Each day, the dispatchers receive a preview of the day’s nominations so they can prepare a plan of system parameters.
The dispatchers are required to meet contract flow rates and pressures at all of the supply and delivery points along the pipeline. “Schematic of the pipeline” Pipeline operation A lot of natural gas pipelines in the U. S. drive in an extremely dynamic fashion. The flows into and out of the system vary during the day. This trend is expected to enhance with the deregulation of the gas industry. This frequently raises the question as to whether there is a single group of set points that will result in optimal operation for an entire day.
Since the introduction of optimizers, the debate of steady-state versus dynamic optimization has raged. It is the view of the authors that steady-state optimization coupled with dynamic online control gives a practical solution to problems in systems which approach steady state most of the time. It is envisioned that considerable changes in the pipeline operations happen once a day. At some time new flow enters the pipeline during the day. This flow may be maintained until the next day. The optimizer has to be run to decide the new operating conditions when these changes are expected.
It may be a good idea to run the optimizer at a higher frequency if the change in flow occurs at a greater frequency. Optimizers by themselves are not sufficient for the dispatcher to run the pipeline in an optimal fashion. The optimization is made automatic by a control scheme, so that the dispatcher can use the control scheme to shift the pipeline from one optimal state to another. Control strategy Most compressor stations consist of local station controllers, which are able to control the suction pressure, the discharge pressure or the flow or some combination of the three.
However, control is localized and the interaction is not taken into account, between compressor stations. In general, during a variation in the flow through the pipeline the controllers interact to a very large extent and can often oppose each other. Model-predictive controllers account for interactions between different control actions and all the control actions are computed to reach a general aim. These controllers also forecast the effect of control action into the future. Hence they can act faster and more efficiently than feedback controllers. Technically, the control scheme has the following three objectives:
1) Keep the pipeline at a most favorable pressure profile (normal mode); 2) Gracefully handle the daily change in nominations (transition mode) with proper unit selection; 3) Automatically handle operations in the event of a compressor going down (event mode). In meeting these control objectives, the overall project objectives are met. The controller maintains the pipeline at a best pressure profile recommended by the optimizer and confirmed by the dispatcher. Pressure profiles at which the compressors are most efficient is chosen by the optimizer. Energy losses can be minimized by this and hence fuel usage.
A scheduling algorithm prepares the pipeline for a change in nominations for the day during the transition mode of operation. The dispatcher with the necessary information to move to the new flow is provided by a heuristic-based scheduling algorithm. On the dispatcher’s approval, this algorithm is used to shift the pipeline to its new set of best possible operating conditions. Units may come up or go down on definite dispatcher commands or depending on some logic in the local control algorithm. The overall control scheme has to be capable to handle both cases.
In the case of local controllers bringing up or shutting down units automatically, the plant control engineers need to be involved in matching the unit selection logic of the local controllers with the logic of the optimizer. An event mode is reached if a compressor goes down creating a shortage of horsepower at a particular station. The downstream stations speed up to account for this shortage. The proposed control scheme is operated by the dispatcher from the central control room. The controllers are in touch with the SCADA and “sit on top” of the local station controllers.
The control scheme gives set points to the local controllers that would then move the system toward optimality. To obtain the optimal pressure profiles and unit selection, the control scheme also communicates with the optimizer. In different industries like ammonia, sulfur recovery plants, and gas plants, the multivariable robust model-predictive control strategy is fairly common. “Schematic of the control system” U. S. natural gas pipeline network The U. S. natural gas pipeline network is a extremely integrated transmission and distribution grid that can transport natural gas to and from nearly any location in the lower 48 States.
The natural gas pipeline grid comprises: 1. More than 210 natural gas pipeline systems. 2. 302,000 miles of interstate and intrastate transmission pipelines 3. More than 1,400 compressor stations that maintain pressure on the natural gas pipeline network and assure continuous forward movement of supplies 4. More than 11,000 delivery points, 5,000 receipt points, and 1,400 interconnection points that provide for the transfer of natural gas throughout the United States 5. 29 hubs or market centers that provide additional interconnections 6. 399 underground natural gas storage facilities
7. 49 locations where natural gas can be imported/exported via pipelines 8. 7 LNG (liquefied natural gas) import facilities and 100 LNG peaking facilities. Interstate – Pipeline systems that cross one or more States Intrastate – Pipelines that operate only within State boundaries Conclusions The control system, along with a commercial pipeline optimizer, controls and optimizes a pipeline segment. The strategy is aimed at minimizing fuel consumption of the compressors, easing the scheduling of changes in gas flow rates, proper unit selection and handling equipment outages.
Commercial optimizers that let the dispatcher run the software are very advantageous because the dispatcher can make decisions based on the optimization results as opposed to individual preferences. The control scheme automates the optimization operation, which keeps the dispatcher free to perform other duties. The advantages of the control scheme should be very apparent in a complicated network, where it is very difficult to predict the effect of changing one station speed on the rest of the system. Knowledge-based scheduling algorithms help considerably in moving the pipeline from one steady state to another.
References Natural Gas Pipeline Network: Changing and Growing. (1999). Retrieved ? November 21? , 2008, from http://www. eia. doe. gov/pub/oil_gas/natural_gas/analysis_publications/natural_gas_1998_issues_trends/pdf/chapter5. pdf Comprehensive Inspection Form – Gas Distribution Pipeline. (2006). Retrieved ? November 21? , 2008, from http://dnr. louisiana. gov/cons/CONSERPI/documents/gasdist/ComprehensiveInspectionForm. pdf Natural Gas Pipelines. Retrieved ? November 21? , 2008, from http://www. nrhtx. com/pdf/Buzz/Natural%20Gas%20Pipelines. pdf U. S. Natural Gas Pipeline and Underground Storage Expansions in 2003.
Retrieved ? November 21? , 2008, from http://www. eia. doe. gov/pub/oil_gas/natural_gas/feature_articles/2004/Pipestor04/ngpipestor04. pdf Russ, D. (2008). Open Season Pathfinder Gas Pipeline Project. Retrieved ? November 21? , 2008, from http://www. transcanada. com/pdf/company/projects/Pathfinder%20Open%20Season%20Procedures%20October%202008. pdf Kirby, S. The Alaska Natural Gas Pipeline Project. Retrieved ? November 21? , 2008, from http://www. iogcc. state. ok. us/Websites/iogcc/Meetings/Midyear_Meeting/Calgary-2008/Sue-Kirby. pdf Basu, U. , Venkataramanan, G. , & Ferber, P. Gas pipeline optimization. RetrievedSample Essay of Paperial.com