Sidney P. Santos, Petrobras Gas & Energy, Rio de Janeiro, Brazil;

and Saeid Mokhatab, Independent Consultant, Iran

Compressor station design has been attracting a growing amount of attention in recent years, since it plays a very important role on the feasibility of gas pipeline projects. However, the decision on compressors arrangement, whether series or parallel, is mostly based on economics and failure analysis.

This type of analysis is useful to evaluate capacity reduction due to compressor units outages and also to support the decision on the installation of standby units, while considering exposure to potential penalties from transportation contracts. The goal here is to emphasize the need to simulate different sizes and arrangements for compressor station units, and the potential economic risk of selecting an off-the-shelf solution that might not be the best in terms of capital investment, fuel usage and optimal energy management.

As an example, the Bolivia-Brazil gas pipeline project shows how this type of analysis can enable designers to optimize compressor station arrangement and improve their understanding of compressor station operation. It also provides informed guidance that will enable designers to identify the best alternative with regard to the sizing and arrangement of units for capacity expansion.

System configuration

With the Bolivia-Brazil gas pipeline, natural gas is taken from proved and probable Bolivia reserves of around 23.37 Tcf (0.66 trillions of cubic meters) to the Brazilian market. Its northern leg runs for 1,127 miles with 32-in. diameter pipe, from Rio Grande, Bolivia, up to the city of Campinas in the state of São Paulo, Brazil. The southern leg runs for 683.5 miles with 16 to 24-in. pipe, from Campinas to the city of Porto Alegre in the south of Brazil.

The Bolivia-Brazil gas pipeline interconnects to the existing gas pipelines in the southwest states of Brazil, and becomes a gas transmission network (Figure 1) with four sources of natural gas. The gas importation contracts from Bolivia totals 1,043 MMscfd (30 MMcmd); from Campos Basin varies from 278 to 625.7 MMscfd (8 to 18 MMcmd); and from Santos Basin varies from 41.7 to 62.5 MMscfd (1.2 to 1.8 MMcmd).

Design considerations

Compressor stations have to be considered as an integral part of a gas transmission system, and the design of the system has to consider all possibilities of incorporating compression as the system develops and throughput is increased. A given system may have anywhere from a few stations up to over 50. These stations add enough energy to the gas to overcome frictional losses and to maintain required delivery pressures and flows.

The question of whether a station should be equipped with compressor units in series or in parallel cannot be answered universally. While a series arrangement may present some advantages when standby compressor units are not required, parallel arrangement provides better results when standby units are required. Parallel arrangement also provides more operating flexibility under failure scenario analysis and are better suitable when capacity ramp up has to be considered to address market needs.

A pipeline designer should evaluate different configuration and compressor unit size, and perform technical and economical feasibility studies to identify which configuration presents a better overall economic result. The decision process has to take into account issues such as capacity ramp-up, further expansion, back-up strategies, operational strategy and transient analysis.

Transient simulation

Transient simulation, including failure analysis for compressor station units (with the operating conditions presented in Table 1), is fundamental for the best selection of compressor unit and configuration. This type of analysis best fulfills the technical and economical requirements and helps designers to make informed decisions regarding equipment availability and reliability.

A number of studies were conducted for the original design of the Bolivia-Brazil pipeline. These included technical and economical analysis, thermo-hydraulic studies, and failure analysis. These studies recommended a parallel arrangement for the fourteen compressor stations, with four 7,000-hp ISO gas turbines per station. The project schedule considered the installation of three initial compressor stations that were necessary for three years of operation. These studies evaluated bigger turbocompressors, with series and parallel arrangement. They also evaluated electric-driven compressors as an alternative, since there is an electric high-tension transmission line parallel to the gas pipeline on the Brazilian side.

With more accurate information from natural gas market development, a new set of studies were performed on contractual capacity, ramp up-related changes, and experience acquired from the original design.

Design approach

The procedure adopted to select the compressor station units and configuration follows a sequence of steps to save engineering and simulation time:

Step 1. Run steady and transient states for the model with generic compressor and driver at each new compressor station and for each operation year. Set a maximum power available for maximum capacity at each station and compare with market availability in terms of compressor and driver. In Figure 2, one can see the effect of putting a limit on the maximum power available. The station discharge pressure drops when maximum power is reached and more capacity is demanded from the system.

Step 2. With the adjusted power profile and compression ratio taken from the maximum capacity transient run, a pre-selection of compressor and driver was made based on manufacturers availability. A transient analysis with the pre-selected equipment was performed and adjustments were discussed with manufacturers so as to guarantee the selection process.

Step 3. With the equipment pre-selected and adjusted for maximum capacity, the operational history was checked in terms of installed unit requirements and necessary changes in compressor impellers.

Step 4. Series and parallel arrangement were then checked for fuel usage and failure analysis.

Since the Brazilian side of the pipeline has a high tension line running in parallel to the pipeline (Figure 3), the alternative using variable-speed electric motor drive was considered for both parallel and series arrangement for the stations.

Considering the above-mentioned materials, simulation analysis was made for the following alternatives:

  1. Three turbocompressor units per station in parallel arrangement
  2. Three electric-driven compressor units in parallel arrangement
  3. Two electric-driven compressor units in series.

The electric drivers presented the characteristic of being modulated to the station power requirement and together with electric power availability became an attractive option.

Pre-selection of equipment

To allow pre-selection of equipment, a text file from the transient run for each year of operation (including the variables necessary to size centrifugal compressor and drivers) was sent to the manufacturers. This approach proved efficient, and led to detailed technical discussions between design engineers and manufacturer representatives. These meetings resulted in a good understanding of project needs and requirements. Some (although not all) pre-selected equipment is presented in the following paragraphs.

Alternative A. Three turbocompressors per station in parallel:

Pipeline maximum capacity: 1.1819 Bcfd (34 MMcmd)

Compressor type: Solar C452 C2F-B3R

Gas turbine: Solar Mars 100 – 15000 TMF-2

Alternative B. Three electric-driven compressors per station in parallel:

Pipeline maximum capacity: 1.2167 Bcfd (35 MMcmd)

Compressor type: Solar C452 C2F-B3R

Electric motor (variable speed driver): Siemens – 10 MW

Alternative C. Two electric-driven compressors per station in series:

Pipeline maximum capacity: 1.2167 Bcfd (35 MMcmd)

Compressor type: Cooper RFA36

Electric motor (variable speed driver): Siemens – 16 MW.

Performance results

All the alternatives presented above were simulated in steady and transient states. Compressors and drivers where selected to optimize each alternative. Additionally, a failure analysis was performed in transient state to identify the alternative with better operational performance.

The delivery pressure at the downstream end of the trunkline (a refinery in Campinas) was plotted to measure the failure effect (Figure 4). The effect on delivery pressure (time dependant pressure reduction) obtained from transient compressor unit failure analysis is a very important aspect to be tracked. The alternatives were ranked based on the delivery pressure effect. A station designer must pay special attention to the delivery pressures at the ends of the pipeline network under analysis. The feasibility analysis identified Alternative C (series arrangement–two units) as the best one; Alternative B (parallel arrangement-three electric-driven compressors) as the second best; and Alternative A (parallel arrangement-three gas turbine-driven compressors) as the third best.

Capacity comparison

Thermo-hydraulic simulation was performed for each predicted operation year to identify the gas pipeline transmission capacity. Figure 5 presents a comparison graph for each alternative. It is clearly evident that the alternatives with electric-driven compressors provide higher capacities than gas-turbine driven compressors, since fuel gas is not required.

Conclusion

The basic conclusion from these studies was that we cannot overlook the importance of steady-state and transient simulation as an optimization tool for a gas pipeline and compressor station design. The most appropriate selection of compressor units and arrangement requires transient analysis, regardless of whether the operator is considering series or parallel units.

Although some might consider this approach to be too time consuming, experience argues otherwise. Every bit of effort performed in the design phase will help define the technical, operational and economical aspects of a project, and having this information will save a lot of time during implementation. It is also important to define the key pipeline operational parameters that will be included in the transportation contracts. The primary conclusion here is that compressor unit sizing, arrangement definition and driver selection need to be part of a technical and economic study that will identify which alternative is the best one on a tailor-made design, and not based on general rules.

Acknowledgment

Based on a paper presented at the 32nd Annual PSIG meeting held in Savannah, Georgia - USA.