Evaluating Overpressure
One of the greatest challenges we face in conducting PHA analysis is properly evaluating the consequences of our PHA deviations. Today we want to examine the factors to determine overpressure consequences and severities. Typically we must answer the following questions.
How far will the overpressure consequence escalate - no consequence, minor loss of containment or will a major catastrophic event involving rupture, severe harm, fire, and explosion occur?
What are the factors and criteria that need to be identified to accurately determine overpressure consequences?
Lets consider the following HAZOP example scenario in our discussion. Perhaps you or your company have found yourself in this situation?
A pressure vessel (V-100) facilitates separation of LPG (propane/butane). We discover a blocked flow cause for overpressure from a gas outlet pressure control valve (PCV-100) being inadvertently closed. The following conditions and categories of information were used to form the basis for making decisions.
Operating Conditions:
The product is found to contain 1000 ppm H2S. IDLH (Immediately Dangerous to Life & Health) is 100 ppm. H2S levels are expected to be below 100 ppm within 1 meter of the vessel with occurrence of minor leaks. A major rupture could result in toxic levels of a few hundred ppm resulting in serious injury to personnel within the direct path of the release.
The product may ignite if a sufficient release volume and velocity occurs to create a vapor cloud and static ignition within the explosive limit. Otherwise minor leaks will result in LPG levels being below the LEL (Lower Explosive Limit). There are no ignition sources within 100 meters of the vessel and the area is not accessed by vehicle and rarely visited by personnel. The site has frequent prevailing winds, but there are daily and seasonally times where atmospheric conditions could create increased gas concentrations.
The upstream process pressure operates at 400 psig. The vessel is subject to daily and seasonal pressure and temperature cycling.
There have been 3 historical overpressure pressure incidents from PCV-100 control valve failing closed. The PSV successfully operated and relieved the overpressure at MAWP (Maximum Allowable Working Pressure) to the flare system in each case. Operations was not able to intervene quickly enough to prevent overpressure, even with activation of the high pressure alarm.
The PSV is inspected and tested every 4 years and has failed its pre-pop test once (relieved at 1.4 X MAWP or 350 psig) in a total of 4 turnaround cycles due to improper repair from a third party vendor. The company has no external QMP (Quality Management Program) inspection program.
No SIS (Safety Instrumented System) high pressure shutdown control or inlet shut off valve or pressure relief system, exists on the vessel. No previous accurate risk studies or additional safeguards were completed to prevent overpressure re-occurrence, which is estimated to occur on average every 5 years.
2. Design Conditions:
Integrity, Reliability and Maintenance engineering were contacted to confirm manufacturing design and confirm overpressure consequences.
The vessel was built in 2000 under revised ASME codes. DP (Design Pressure) and MAWP of the vessel and its piping is 250 psig. The hydrotest pressure was 1.5 X MAWP or 375 psig. The design overpressure safety factor is 3.5 X MAWP or 875 psig.
The vessel metallurgy is carbon steel. Its yield point (elastic limit) is determined to be 1.9 X DP or 425 psig.
3. Maintenance History:
The vessel undergoes cleaning and inspection every 4 years. The following was found and completed in its 22 year history
4. Industry, Regulatory Standards and Company guidelines:
Common industry standards were found to adopt a very conservative approach and state that the risk of vessel rupture and subsequent catastrophic consequences are probable, after the pressure exceeds 1.1 to 1.3 times the DP or MAWP. The standards list no other criteria for for determining overpressure consequences, other than assuming that the equipment is operated, maintained, and designed to industry and engineering standards (as per RAGAGEP, Recognized and Generally Accepted Good Engineering Practices). However, the standard does provide allowance for site specific analysis and data to be utilized in consequence analysis but state that most companies do not have the required level of information and processes in place to validate vessel integrity and acceptance of higher overpressure exposures.
The regulations and regulator do not utilize a process safety risk based approach in determining overpressure consequences. They stipulate the vessel must be operated, designed and maintained to engineering standards (ASME, B51/52, API 510 for example). All overpressure incidents must be reported to the regulator within 24 hours, be thoroughly investigated by the chief pressure equipment engineer of the company and regulator, and mitigations must be implemented to prevent re-occurrence. The regulator reserves the right to conduct further inspection and halt operation.
The company is finding several overpressure scenarios while conducting their PHA studies and this is generating significant internal as well as external PHA consultant debate, especially as some are wanting the adoption of the conservative industry standard, while others are claiming that they have never experienced these consequences in their operating history, even though they are possible. The cost of implementing additional safeguards is significant and the company has limited resources due to internal and external business conditions. No additional overpressure consequence analysis and metrics are available or have been conducted in the company's operation. The company's process safety representative is seeking to develop site specific overpressure guidelines for conducting HAZOP and PHA studies, to accurately determine the risk using internal resources and to convince management to install required safeguards to reduce the risk to tolerable levels.
After carefully analyzing the overpressure consequence and associated conditions, the PHA team made the following conclusions.
If PCV-100 remains closed then V-100 LPG separator pressure would rise to the maximum upstream process pressure of 400 psig or 1.6 X its MAWP/DP of 250 psig. Although the vessel pressure would exceed its MAWP/DP and HTP, the yield point of the metallurgy would not be reached, so no elastic deformation, and rupture is not expected.
However, because the inspection and reliability program maturity has not been fully developed, and the overpressure incidents have not been fully resolved, it was determined that leaks would occur at the flange joints, gaskets, fittings and potentially compromised areas in the vessel.
These leaks, under certain atmospheric conditions and personnel presence, could lead to potential toxic H2S exposure levels and fires. The PHA team decided that serious permanent injury, financial, and business and reputation consequence severities were credible.
The PHA team determined that the frequency of the cause (PCV-100 failing closed) was on average every 5 years, but further research revealed that 2 events occurred within 11 months. Therefore a more conservative IEF (Initiating Event Frequency) of once per year was adopted instead of the industry standard 1 in 10 years, based on their own facility operating history.
The PHA team determined that a severe inherent risk existed for the scenario, before safeguards, modifiers or enabling events were applied.
When reviewing the safeguards/IPLs the following was determined:
After careful examination of the consequence scenario with scrutiny of the design and operating conditions, discovery of the increased IEF and proper evaluation of existing safeguards, the PHA team determined that a severe residual operating HSE risk existed. The consequence was successfully analyzed and recorded in the HAZOP360 software as follows. The company's risk criterion is that "all residual risk must be immediately reduced below their "Tolerable Event Frequency (TEF)" of 1 X 10-4 for a fatality to continue operations. Therefore, interim control and operating recommendations were immediately implemented before more permanent measures were put in place.
The following recommendations were issued:
Implement immediate measures to further control process flows and vessel pressure to allow operators sufficient time to prevent overpressure. Suggested solutions include:
Remove and test the PSV to verify its function and implement long term PSV maintenance and performance standards to meet 99% reliability.
Develop a permanent long term pressure or flow control/alarm solution to prevent overpressure from occurring. Suggestions include:
Consider installing fixed gas detection at site to detect a loss of containment, and develop remote isolation procedures to prevent personal entry
Perform complete reliability inspection of the entire vessel, and all of its process inlet and outlet piping, flanges and fittings to ensure its full integrity. Ensure a long term pressure equipment integrity program is established with a reliability performance standard of 99.9%
Ensure low point draining, and heat tracing reliability is fully performed and verified. Establish a winterization and dead leg program with a 99.9% reliability performance standard
In conclusion, restoration of the PSV reliability and verification of the pressure equipment integrity will successfully mitigate an overpressure event and prevent an uncontrolled release. However, developing further process flow and pressure controls to allow adequate operator intervention and/or prevent overpressure escalation is always preferred. It is important to understand and research the unique process control philosophy, operating history, equipment integrity and safeguarding reliability when analyzing overpressure consequences.
Do you have any final comments and conclusions on this overpressure scenario example and the factors considered? Do you think this situation realistically exists in industry? What has been your experience? Do you have any comments on how the consequence analysis was developed in HAZOP360. We look forward for your valuable insights and comments.