Mar-2005

Using process simulators will make your plant more productive and efficient

The use of process simulators to model plant operations can provide a plant hundreds of thousands of dollars each year in increased production and lower energy costs

W G “Trey” Brown, Bryan Research & Engineering
Leonard Ochs & Williams J W Varner, Quicksilver Resources

Article Summary

This paper looks at several example plants where process simulators were utilised to optimise their operation and measurable results were obtained. Each of these plants were able to improve their bottom line profit because a process simulator was available and plant personnel were dedicated to using it to improve plant performance and efficiency.

In today’s economic roller coaster, where product margins can be positive one day 
and negative the next, a plant must be designed and operated with the utmost operating 
flexibility, while maintaining high energy efficiency. The process simulator allows both the designer and the operator to maximise this flexibility and determine the best way to operate the plant at both ends of the operating spectrum. Today, a plant that does not use a simulator to monitor its operation is simply throwing money away.

Introduction
Process simulators have been in widespread use throughout much of the energy industry for almost 20 years, and progress continues to make these programs faster, easier and more accurate in the design and operation of processing plants. Process simulators have become required engineering tools for design engineers, 
process engineers and plant operators. When properly used, the simulator can aid the engineer in becoming more productive and creative in performing his job responsibilities. However, many of today’s plants are designed without utilising the simulator’s full capabilities to help optimise the plant configuration and overall cost. Additionally, many operating plants have not been simulated since they were first brought on-line, leading to inefficient operation and uncertainty about how changes in one area of the operation affect other areas of the plant. Understanding how your process simulator works, utilising all its capabilities and being able to compare simulation results to actual operating conditions will result in a better designed and better operated process plant. Process simulators can and should be used in numerous ways, from improving initial plant designs to increasing operating efficiency and/or product recoveries to performance of parametric studies to determine the effect on plant operation at varying conditions. This paper concentrates on examples of how simulators were used in the design and operation of four different plants and the resulting benefits that were derived in each case.

Plant A
This plant was originally designed using a process model taken from an identical plant, operating in parallel to the new plant (Figure 1), and modifying the simulation as required to meet several new design parameters. In addition, the simulation was expanded to include both plants, so that the operators could predict how a change in one plant’s operation would affect the other. During the design phase, it was determined that, by sharing the excess residue compression capacity of the new plant with that of the existing plant, the existing plant’s inlet rate could be increased by 10%, while at the same time increasing product recoveries (see Table 1). Thus, by using the simulator to model both plants in parallel operation, the designer was able to provide the operator with a feature that would increase plant throughput by an additional 20 MMscfd and total revenue by over $30 million per year (using $4.50/MMBtu) at a cost of less than $300 000! In this case, use of the simulator and the ability of the designer to fully utilise it, resulted in a tremendous benefit to the operator, both monetarily and by strengthening his processing position in the region.

During start-up of this plant several common difficulties were encountered. These included compressor start-up problems, product pump problems and loss of the inlet gas treating system. These problems resulted in plant upsets that, at the time, seemed normal to all plant start-ups. However, after being in operation for approximately two months, a plant performance test was run and the plant did not meet the guaranteed product recovery levels (see Table 2).

Simulations were performed on the plant, using actual operating data, and it became apparent that something was wrong with the demethaniser. The simulation could be “forced” to match the actual operating data almost exactly, but to do this required using tray efficiencies of only 20-25% compared to a typical 60-65% tray efficiency for cryogenic demethanisers (see Figures 2A and 2B). Additional tests were performed, including a gamma-ray scan of the tower, and it appeared that tray damage had occurred. The general consensus was that the tray manways had been dislodged from the main trays by pressure excursions in the tower during one or more of the plant upsets. When the column was opened, this was indeed found to be the cause of the problems. Once fixed, the plant was put back on-line and performed extremely close to the design simulation model predictions, exceeding product guarantee levels by 3%. Therefore, while the simulator did not specifically predict what the problem was, it did aid those who used it in determining potential causes for the problem and gave them a path to pursue for further confirmation.

Plant B
This plant is actually a compilation of several plants, all with various plant processes and operating both in parallel and interwoven with one another (see Figure 3).

As in Plant A above, the designer had to develop an overall plant model that would accurately depict how all of the plants would interact and the impact a change to one would have on another. Intermingling of plant amine systems and hot oil systems led to higher energy efficiency and lower overall capital costs for the plant. Utilising a common residue gas header system allowed the operator to take full advantage of his existing residue gas compression and handle swings in flow and discharge pressure, yet avoid operation of excess compression, thus saving fuel and maintenance costs. Accurate prediction of the different system interactions and determination of how to operate in the most efficient manner possible would have been very difficult and time consuming without a process simulator.

Several opportunities for improving the efficiency of this plant have arisen over the years and, in each case, a simulator has helped in identifying these areas of improvement. The following lists a few examples of these. On one occasion, it was determined that one demethaniser was not performing in the manner it should. When the actual temperature profile across the column was inserted into the process simulation, it showed that the side reboiler was providing approximately 95% of the reboiling duty to the tower, while the bottom reboiler was having minimal input (see Figure 4A). The design simulation showed that the side reboiler should have only provided two-thirds of the column reboiling duty and, operating in the manner it was, approximately 25% of the column was totally ineffective in the fractionation process. This lowered product recoveries and made it difficult to meet product specifications. Upon inspection, it was noted that the liquid draw to the bottom reboiler had been almost completely “pinched off”. This resulted in the side reboiler being supplied with warmer inlet gas as the heat medium and thus providing more heat transfer. The bottom reboiler draw was reopened and the side reboiler draw was pinched back to some degree. Within two hours, the tower temperature profile had been re-established to a point that was very close to the simulation model and product recovery levels reached predicted levels, without encroaching upon specification limits (see Figure 4B).