Read part two of this article here.
Dynamic HAZOP analysis
Over the years, dynamic analyses have been performed to support the internal HAZOP reviews. Most of these are very detailed and require specific user information such as ‘as built’ equipment and line size data as well as detailed valve performance data. Other flowsheets are more simply configured and are used occasionally in HAZOP analysis to confirm and/or support the qualitative conclusions normally encountered during the reviews. The following examples illustrate some process scenarios typically encountered during HAZOP reviews.
Conventional column: loss of reflux
The majority of process units utilise two and three phase conventional separation columns. These columns typically have an overhead condenser with a receiver and a bottoms re-boiler circuit.
In a typical HAZOP analysis, there could be potential hazards associated with the installation of engineer controlled emergency isolation valves. In those situations, an evaluation will be conducted to determine whether the installation of the engineer controlled emergency isolation valves provides the safer option for personnel, the environment or surrounding equipment than if these valves were not installed. These isolation valves can be local or remote manually operated and are typically not interlocked to other instrumented devices.
From the database of UNISIM® templates, a dynamic model of the conventional column follows a standard process flow diagram view of equipment, piping and basic piping and instrument diagram control designations. This model has event schedules preconfigured with a number of upset scenarios. If needed, adhoc scenarios can easily be added to the model during the HAZOP team analysis.
To simulate an upset event, a manually operated isolation valve is inadvertently closed during normal operation. First, the column is lined out to steady state conditions; then the isolation valve is triggered to fully close. Through a sequence of events, the overhead vapour from the column immediately spikes upward, then moves downward rapidly as it is displaced by the increasing liquid level in the receiver. After 3.5 mins the liquid level completely fills the receiver and overflows through the vapour line at a large rate. With no reflux flow returning to the column, the liquid level in the sump starts to decrease. The liquid level ultimately drops to zero after another 7.5 mins. During this upset the column operating pressure increased close to the pressure safety valve set pressure but did not open.
Obviously, during normal operation, the appropriate trip activation or operation actions in response to safety alarms would have been in place to mitigate any severe operating scenarios such as the one performed in this simple example. As this example shows, the continued use of dynamic simulation in the HAZOP reviews has allowed for quantitative evaluation of the effect of time on process safety issues, particularly as they relate to equipment failures, startup and shutdown scenarios. Vendor data can also be readily incorporated to validate performance. For instance, the actual valve flow coefficients CV and stroke time data is used for all control and isolation valves whenever dynamic simulation work is being implemented in HAZOP analysis.
Phenol process: operating pressure evaluation
Dynamic studies are also performed for users based on internal HAZOP reviews. One user requested the possibility of running oxidiser tanks at higher pressures, so as to increase product flow, should be evaluated.
The process flow diagrams for the oxidiser section were taken from a library of design templates and modified to include all user specific information. Piping and instrument diagrams were also used to accurately define the placement of all control and isolation valves as well as control logic systems in the dynamic simulation. Some elevations and line sizing information were also extracted from these drawings. The remaining data was obtained from previous hydraulic reports. Equipment sizes were also incorporated into the model.
The air feed valve/isolation valve configurations utilised the ‘as built’ detailed valve performance data provided by the user.
The results of the dynamic analysis showed that the user could run their unit at higher operating pressures. A greater pressure trip point was recommended for the oxidiser shutdown system, as this would still prevent overpressure of the oxidiser tanks from reaching their set relief pressures. Based on the ‘as built’ stroke time valve data obtained from the customer, it is clear that the valves can perform quickly in avoiding pressure relief of the oxidiser tanks at the highest possible operating pressure.
The combination of traditional HAZOP analysis with dynamic simulation facilitates accurate determination of the consequences of deviations from normal design. Potential hazards and safety issues can be better identified. By dividing the problem model into smaller, independent flowsheet sections, the level of complexity associated with the dynamic simulations can be reduced.
HAZOP teams can now test the effectiveness of mitigation strategies for many emergency situations. The use of dynamic simulation as a process safety/risk assessment tool provides significant added value to HAZOP reviews and reports.
Written by Scott M. Wozniak and Bill Weide, UOP, a Honeywell Company, USA. This is an abridged article taken from Hydrocarbon Engineering’s January 2016 issue.
This article is based on a paper that was first presented at the American Institute of Chemical Engineers' 2015 Spring Meeting and 11th Global Congress on Process Safety, held from 27 - 29 April 2015 in Austin, Texas.
Read the article online at: https://www.hydrocarbonengineering.com/special-reports/22012016/taking-safety-to-the-next-level-part-three-2182/