Using dynamic analysis to reduce the weight of offshore installations
Jeffrey Feng, Avanish Aggarwal, Surajit Dasgupta, Hoss Shariat - KBR
Minimizing equipment weight and dimensions reduces the installed cost of an offshore processing facility. Because equipment specifications must consider non-steady state operations, offshore engineering contractors apply dynamic analysis to evaluate equipment design for initial startup, normal operation, turndown, shutdown, restart, and variation in plant throughout. The knowledge gained from dynamic analysis is used to minimize over-design, explore alternative equipment with a smaller footprint, and ultimately achieve a design that is most appropriate for the offshore environment.
The scope and emphasis of dynamic analysis evolves as the project moves forward. To fully appreciate the extent of knowledge that can be gained from dynamic analysis, examples from past offshore installations are given below and cover the three main applications: flowsheet analysis, material selection, and equipment detailed design.
Flowsheet analysis
A project typically starts with a feasibility study to develop the overall flowsheet. Minimizing equipment count during the feasibility phase can improve the overall cost matrix of a project, particularly for brownfield (material field) development where the addition of equipment must be kept to a minimum. For brownfield development, dynamic analysis verifies the operating limits of existing equipment to determine if it can be reused.
Recently, the developer of an oil field planned to modify an offshore platform to accommodate the changes in reservoir conditions and the tie in of new flowlines. The original platform had an export gas compressor with an inlet cooler that also functioned as a recycle cooler. As part of the brownfield development, a new compressor was to be added. However, the developer wanted to retain the existing exchanger as a common recycle cooler for both compressors, and requested a feasibility study for this somewhat unusual configuration.
Because the exchanger was to be used as an aftercooler in normal operation and as a recycle cooler in startup and shutdown, a dynamic analysis was performed to fully understand the exchanger load for both steady state and transient operations, and to determine whether the existing exchanger could accommodate the predicted load for the entire range of operations with the new compressor.
To ensure accuracy of the dynamic analysis, first a simulation model was built for the existing installation and validated against the steady state operating data to ensure the accuracy of compressor curves, heat transfer coefficient, valve capacity, and other equipment parameters in the model.
After validation, a new compressor block and its recycle line were added to the model. A startup simulation was then performed on the model.
The maximum exchanger duty occurred when the compressors were operating in total recycle just before the discharge pressure reached the export pipeline pressure, i.e. point “D” on the graph below. The predicted flowrate, temperature, and pressure corresponding to the maximum load were extracted from the simulation model, and a detailed rating calculation for the exchanger was performed. It was concluded that the existing exchanger could meet the cooling demand for both compressors and a new exchanger would not be needed.
This example shows that dynamic analysis can be used early in the feasibility phase to develop an optimum design with the least impact on an existing facility. Generally, this type of analysis is done for brownfield development projects in the North Sea where the stakeholders are considering a cost-effective way to accommodate the new reservoir conditions after the facilities have been in continuous operation for over 10 years. For greenfield (new field) projects, dynamic analysis is infrequent during the initial feasibility phase due to a lack of data.
Material selection
When a greenfield project moves beyond the initial feasibility phase, a front-end engineering study usually is performed by an engineering contractor, with opportunities to select the appropriate material for construction considering factors such as resistance to corrosion, weight, cost, etc. Examples are the cooling water and fire water systems on an offshore platform. These systems generally have large diameter pipes that run through the entire facilities. For these systems, use of fiberglass reinforced plastic (FRP) pipes can save a lot of weight compared to metallic pipes. However, excessive dynamic force can be detrimental to FRP pipes. The most severe dynamic forces generally occur during pump start and pump trip when there is a rapid change of fluid pressure and velocity. To ensure the FRP pipes are specified properly, dynamic analysis determines the maximum and minimum pressures, maximum velocity, and dynamic forces for each pipe segment in the systems.
Predicted exchanger duty for startup. Point A: Initial condition, with both compressors idle; Point B: The compressors were accelerated to the minimum governing speed; Point C: After the compressors reached the minimum speed, the compressor controllers were switched to automatic, which partially closed therecycle valves; Point D: the compressor speed was ramped up until the discharge pressure reached the export pipeline pressure, which started a forward flow through the compressors; Point E: the compressors reached the normal operating points with the recycle valves fully closed.
Below is a diagram of a seawater system, comprised of several lift pumps, multiple exchangers in parallel, and large diameter pipes. This system had considerable elevation change, with the pumps located below sea level, but some of the exchangers were on the upper deck of the topsides. This large elevation difference made this system susceptible to severe waterhammer if one pump was tripped and column separation occurred at elevated locations.
To fully evaluate the operating envelop of the system, simulations were performed for pump start, trip, and restart. Different design configurations also were tested. The figure below shows that the pressure transient at the pump discharge manifold could be reduced by adding an 18-in. (20-cm) pump bypass line with a dump valve. With the bypass line, maximum and minimum pressures were within the allowable limits specified by the FRP vendors, so FRP pipes could be used.
In this facility, the use of FRP in the cooling water and fire water systems reduced the piping weight by over 50%. This example is based on a recent project offshore Southeast Asia, and is typical of new deepwater installations around the world where weight management becomes more important.
Detailed equipment design
When a project moves to the final detailed engineering phase, the overall flowsheet and equipment selection will have been fixed, but equipment weight still can grow. Often, the project developers already have established an upper limit for the weight and dimension of the topsides, which challenges the engineering contractor to reduce the equipment weight and dimension without comprising the safety and capacity of the facility.
One area that dynamic analysis has been used to reduce equipment weight and dimension is large shell-and-tube exchangers in the gas compression system. The following diagram is a schematic of a typical offshore facility with gas/oil/water separation trains and gas compression trains. The shell-and-tube exchangers shown are operated with low-pressure cooling water on the shell side, and high-pressure process gas on the tube side.
In the event of a tube rupture, high-pressure gas will escape from the ruptured tube and pressurize the exchanger shell. Therefore, the shell of these exchangers in offshore normally is designed for instantaneous tube rupture. Dynamic analysis has been used to design a surge relief device on the exchanger. Such analysis is particularly useful in an offshore facility where the gas in the tubeside is generally at high pressure, as shown in the following example of an exchanger on a floating production platform:
- Maximum gas pressure of 230 barg in the tube
- Normal operating pressure of 3 barg in the shell
- Design pressure 10 barg for the cooling water piping adjacent to the exchanger.
The outside diameter of the tubes is 0.75 in. (1.9 cm). The exchanger shell is 5 m (16 ft) long and 1 m (3 ft) in diameter. The exchanger is DFU type and has a horizontal baffle. In the original design, the maximum allowable working pressure of the shell was 10 barg, and three relief valves were connected to the shell for overpressure protection -- one near the inlet water nozzle, one near the outlet water nozzle, and one near the elliptical end. Because the exchanger was on the lower deck of the platform, this configuration required three long 6-in. (15-cm) relief lines to connect the exchanger to the relief valves and relief header on the upper deck.
To minimize the piping for the exchanger, two alternatives were proposed:
1. Design the shell to 153 barg, or two-thirds of the maximum tubeside pressure. There would be no need to install any relief device to protect the exchanger shell, but the exchanger would be about 20% heavier than the original design.
2. Design the shell to a pressure higher than 10 barg but below 153 barg, and install a rupture disk to protect the shell. To improve the effectiveness for overpressure protection, rupture disk instead of relief valve was used, and it was to be mounted directly on the shell.
For the second option, a dynamic analysis was performed to determine the shell design pressure with only one rupture disk. The figure below shows the results of a simulation case where the tube rupture was assumed to occur near the inlet water nozzle (Point A). The maximum pressure that was applied to the shell was about 30 barg, and after discussions with the project developer, the shell design pressure was changed to 20 barg. The adjacent cooling water lines were also analyzed to confirm that the change in relief configuration did not create overpressure problem in the pipeline system.
The increase of shell design pressure from 10 to 20 barg eliminated two long vertical pipes from the platform and reduced the weight of the exchanger package after factoring in fewer rupture disks and relief headers. This example is based on a deepwater project in the North Atlantic, and is typical of offshore processing facilities with high-pressure gas injection.
In addition to optimizing exchanger weight, dynamic analysis also can minimize equipment over-design. This type of analysis has been done for onshore installations including flare and utility systems, but the benefits to offshore installations can be even greater. Opportunities for reducing equipment footprints tend to be more abundant in the detailed engineering phases as compared with the previous feasibility and front-end engineering phases, and it is becoming standard for dynamic analysis to be used in detailed engineering for offshore projects.
Model validation
To ensure the accuracy of dynamic analysis, the model should be validated for both steady state and transient operations. Validation of the model against steady state operation ensures that the dynamic model uses accurate equipment data such as exchanger heat transfer coefficients and compressor operating curves, which are important for brownfield analysis where operating conditions may have changed over the years. After the model is validated in steady state, the next step is to validate the model against transient operations such as startup and shutdown. A properly built dynamic model can reproduce field measurements over a wide range of operating conditions. As an example, the discharge pressure of a two-stage centrifugal compressor at an existing offshore facility was measured during startup and compared against the model prediction.
Transient operations are the controling design cases for many equipment units in offshore installations. It is only through dynamic analysis that the operating conditions in transient operations can be analyzed and an appropriate design with the least equipment weight and dimension can be achieved.
Author’s note
The dynamic analysis presented in this article used first-principle dynamic software, including Hysys from Aspen Technology and PIPENET from Sunrise System.
Acknowledgment
The authors thank members of KBR’s Publications Committee for their technical review and insights, as well as many individuals within the company whose contributions in various dynamic simulation projects made this paper possible.
References
Simpson, I. I., “Tubing Rupture in Liquid-Filled Exchangers,” AIChE Loss Prevention Symposium, Volume 6, 1972, pp 93 to 98.
Cassata, J. R., Feng, J., Dasgupta, S., and Samways, R., “Prevent Overpressure Failure on Heat Exchangers,” Hydrocarbon Processing, November 1998, pp 123 to 130.
Cassata, J. R., Dasgupta, S., and Gandhi, S. L., “Modeling of Tower Relief Dynamics”, Hydrocarbon Processing, October 1993, pp. 71-76.
Patel, V., Feng, J., Dasgupta, S., and Kramer, J., “Use of Dynamic Simulation in the Design of Ethylene Plants”, Presentation at the Ethylene Producers’ Conference, AIChE Spring National Meeting, 2008.
Patel, V., Feng, J., Dasgupta, S., Ramdoss, P., and Wu, J., “Application of Dynamic Simulation in the Design, Operation and Troubleshooting of Compressor Systems,” Presented at the 13th Turbomachinery Symposium, 2007.
Wu, J., Feng, J., Dasgupta, S., and Keith, I., “A Realistic Dynamic Modeling Approach to Support LNG Plant Compressor Operation,” LNG Journal, October 2007, pp. 27-30.
