Crude column relief mitigation
Mitigation of plant-wide relief loads can greatly improve the safety of the plant, reduce flaring and be used as the basis for building smaller relief systems
Viewed : 7404
Crude distillation columns â€¨and other main fractionators within a refinery are typically major contributors to a refinery relief load. This is especially true in site-wide power failure cases, where multiple units are being relieved simultaneously. Large relief loads require a large capital investment in the relief system, safety and environmental concerns during the relief event, and loss of valuable hydrocarbons.
Mitigation measures can be used to reduce these relief loads. With proper high integrity instrumentation systems, they can also be used to reduce capital investment related to adding or expanding relief systems. Implementing mitigation measures, however, requires diligence in the analysis of the system, design of the mitigation methods and regular maintenance to ensure proper reliability.
This article focuses on operational mitigation measures for crude columns in a site-wide power failure scenario. A case study is used to test the effectiveness of these mitigation measures and to present the expected relief load reductions.
Process description of relief scenario
Figure 1 shows a schematic of the crude column being studied. During a site-wide power failure, the crude feed and flashed gas to the column stop, initiating the bottoms recycle stream to open. (Note that on feed loss the bottoms stream is recycled back to the heater inlet using a turbine-driven pump. This setup is used to protect the heater tubes from overheating.) All pump-around cooling stops, all condenser fans stop and the offgas compressor stops, blocking the vapour outlet.
Mitigation measures can be considered for new or existing installations and are mainly implemented through high integrity instrumentation systems.
For a mitigation measure to be effective, it needs to eliminate or reduce the vapour load and should have a trigger that is measurable, fast acting and reliable. For the site-wide power failure case, loss of power is a good trigger for the mitigation actions. This is because the trigger is measurable and, as the main cause of the relief, it is the earliest event that occurs. Another potential trigger is column high-high pressure. The column high-high pressure signal can be used as a generic trigger. However, by the time the pressure is high enough to trigger the shutdown, it is harder to recover and a relief might still occur.
For the site-wide power failure case, the main contributors to overpressure in the crude column are:
• Vapour generation in the feed heater (due to bottoms circulation, as indicated in the process description)
• Accumulation of stripping steam in the column
• Loss of cooling
• Loss of off-gas flow (due to loss of the off-gas compressor).
Each one of the four contributors to overpressure can be coupled with potential corresponding mitigation action(s), and then the effectiveness, cost and practicality of implementation evaluated to decide the optimum mitigation strategy. For example, the following mitigation actions can prevent or reduce the above contributors respectively:
• Shutdown of fuel to the heater
• Shutdown of stripping steam to the column
• Provide backup power to air coolers, cooling water pumps and pump-around pumps
• Provide backup power to the off-gas compressor.
In the case presented here, the following actions were selected:
• Shutdown of fuel to the heater on loss of power
• Shutdown of stripping steam to the column on loss of power.
These two actions represent a passive-preventative approach, where tight shut-off block valves are used to stop the flow of fuel and stripping steam. In comparison, adding backup power to maintain cooling and/or the off-gas compressor is more of an active response approach, where the pressure is being actively reduced by cooling or direct removal of vapour by the compressor.
The study was conducted using dynamic simulation, since it can accurately predict the transient behaviour occurring during the relief event.
Initially, a base case is established where the mitigation measures are not used, followed by three cases:
• Case 1 with the shutdown of fuel to the heater implemented
• Case 2 with the shutdown of stripping steam to the column implemented
• Case 3 with both measures implemented.
Base case: without mitigation
In the base case, the relief event occurs without mitigation. As explained in the process description, the feed to the column stops and the bottom recirculation valve opens 100%. The liquid inventory in the bottom of the column is gradually vapourised by the large heater duty until the column bottom totally dries out and no more heat is moved into the column from the heater. The size of the column bottom has a direct effect on how long the liquid in the bottom lasts before it totally dries out and the heater duty stops. Two sensitivity cases were used to show the impact of the size of the liquid hold-up at the bottom of the column on the relief. Figure 2 shows the base case duty trends for both liquid hold-up volumes, while Figure 3 shows the corresponding relief trends. In the regular sump case, the liquid is totally vapourised in slightly more than 10 minutes, resulting in a peak relief load of about 440 000 lb/hr at 17 minutes. The delay between heater shutdown and peak relief load is caused by the system capacity as the hot vapours travel through the column and heat the bulk of the liquid from tray to tray. In the large sump case, the liquid is totally vapourised in about 32 minutes and reaches a peak of about 560 000 lb/hr. The peak relief load stayed approximately the same for about 15 minutes (from minute 16 to minute 31).
Add your rating:
Current Rating: 3