Since PSVs are safety-critical devices, considerable care is taken in validation of the PSV capacities. ASME codes provide specific testing requirements to certify the capaci - ties (for example, ASME BPVC XIII Section 7.8). 6 PSV ori- fices are purposely designed as simple nozzles and are well modelled as ideal nozzles together with a capacity correc - tion factor for non-idealities. PSV orifices are selected to ensure they can pass the required flow necessary to prevent overpressure. On the other hand, globe valves used as a PSV bypass are com- monly treated as non-safety-critical devices and not cov- ered under ASME VIII certification. It is critical to note that a PSV bypass may not necessar- ily be sized to pass the same required flow as the PSV. In such a case, even timely opening of the bypass can result in overpressure. The capacity of the globe valve (consider- ing upstream and downstream losses) should be confirmed using vendor-supplied valve Cv. Vendors may only provide the full open rated valve Cv. Therefore, it may be required to assume the bypass must be fully opened when determining the operator response time. It is important to consider the management of change (MOC) for the bypass valves. They are typically considered as ‘bulk’ piping materials, where replacements are selected based on matching the nomi- nal body size and satisfying line-class requirements. This approach does not ensure that the replacement bypass capacity is the same. A note on the P&ID stating the globe valve is safety critical and requirements for the minimum Cv may be appropriate. In the design phase, it is important to align on the pur- pose of the PSV bypass. It should not be assumed that bypass valves are automatically sized for overpressure protection. It is possible they were assigned a nominal size to satisfy maintenance requirements to simply vent the vessel for maintenance purposes and not for overpressure protection; as such, the design may have only considered normal operating pressure upstream of the bypass valve and not design pressure. If the bypass valve is opened when the vessel is close or above design pressure, higher than anticipated flows can result in unexpected vibration or developments of backpressures that exceed the flare lateral design pressure. Reliability Assuming the bypass valve is adequately sized, and it has been determined that the operator has sufficient time to respond, the reliability of the operator to successfully per- form the mitigation should be considered. A PSV being the last line of defence has a high reliability since it must function when all other layers of protection have failed (operator procedures, DCS, SIS). Instrumented systems are not normally credited when determining the required capacity of the PSV, unless a high-integrity-pres- sure-protection system is installed (HIPS/HIPPS). When overpressure protection is provided by other means than a PSV, it is good design practice to ensure it is as reliable as a PSV. For example, the jurisdiction in Alberta (ABSA) requires that alternative means of overpressure protection be as effective as a PSV and be as reliable on
Figure 3 Operator monitoring system
However, many overpressure scenarios will occur more quickly than an operator can respond. For example, con- sider a tube rupture scenario or a blocked-in case resulting from closure of an automatic ESD valve that closes in four seconds. The rate of pressure rise in systems with small volumes or liquid packed systems will occur more quickly. Before planning to use an operator at the PSV bypass, it should first be established: There is clear indication available that the overpressure upset is occurring or imminent, usually via a high-pressure, high-level alarm or equipment run-status alarm. The time measured from recognition via the indicating alarm to opening of the bypass needs to be quantified. Note that additional normally closed gate valves, which also need to be opened on the bypass line, may need consideration. It must be demonstrated that overpressure above code limits will not occur in the time it takes from indication of the upset to opening the bypass valve(s). The operator must be adequately trained to recognise the upset and have a clear procedure to follow – i.e., open bypass and any other valves on the bypass line as required. Bypass valves are typically multiturn globe valves, and it would be required to establish a basis on how many turns of the bypass globe valve can be accomplished per unit time, and how many turns are required to open the valve suffi - ciently. Operator fatigue should also be considered, result- ing from overly long, mostly uneventful periods of being assigned to the dedicated (and admittingly boring) task of being on standby to open a bypass valve (see Figure 3 ). This may result in a higher likelihood of procedural failure or longer response times. If the operator cannot respond in time to prevent the overpressure, this operator intervention would not be con- sidered a credible means of overpressure protection. In the author’s experience, it is unlikely that many typical over- pressure scenarios will allow sufficient time for operator intervention. Bypass valve sizing If it has been proven that the operator has sufficient time to prevent the overpressure by opening the bypass valve, then the next step is to ensure the bypass valve has ade- quate capacity.
58
PTQ Q3 2022
www.digitalrefining.com
Powered by FlippingBook