There is some language in the EPA’s Risk Management Plan program requirements that tends to cause some confusion…
The owner or operator of a stationary source shall revise and update the RMP submitted under ยง68.150 as follows:
(5) Within six months of a change that requires a revised PHA or hazard review
So when are we required to “revise our PHA” other than when we have to do our 5-year revalidation?
EPA states the following in their RMP Q&A (http://goo.gl/DOK7EF)
What constitutes a revision of the PHA?
The rule states that I have to update my RMP whenever I revise a PHA. What constitutes a revised PHA? Every time I go through management of change procedures I make a notation in the PHA file for the process, but would that constitute a revised PHA if the change did not affect the validity of the PHA?
All changes (except replacement in kind) are subject to the management of change of procedures. When processes undergo minor changes (e.g., minor rerouting of a piping run), information is typically added to a PHA file to reflect the change, even though the validity of the PHA is not affected by the modification. These minor changes and the addition of information about the change to the PHA file are not considered a ‘revision’ of the PHA under the part 68. Major changes that invalidate a PHA, leading you to ‘update’ or ‘revalidate’ the PHA so that it accurately reflects the hazards of the process, are considered a revision of the PHA under part 68.
There are countless situations that could drive us to update our PHA(s), but there is one that is 100% certain! Here is the best way I can explain it…
In our PHA we have identified a risk scenario , it’s consequences, and the SAFEGUARDS we have in place to PREVENT the scenario from occurring – PROTECT ASSETS if the consequences occur – and MITIGATE the consequences. If we MAKE A CHANGE to our process that will impact our scenario(s) risks either by INCREASING the PROBABILITY or SEVERITY and/or DECREASING the SAFEGUARDS then we would need to REVISE our PHA regarding the area the change is being made. Here is a scenario to help explain what I am saying:
We have a storage tank that contains a flammable liquid. Our tank has the following PRVENTION SAFEGUARDS:
- an over-flow prevention system with an SIL 3 that includes a level indicator
- a HI Level Alarm (both audible and visual that sound both at the tank and in the control room)
- a HI-HI Level Interlock that will activate the alarm, CLOSE the fill valve, and shut-off the unloading pump
Our storage tank also has the following PROTECTION SAFEGUARDS:
- Secondary Containment
- Tank Farm is remote – normally unoccupied location
- Tank spacing around this bulk tank is 50% greater than required
- Fixed Foam System within the dike
- Fixed water monitors on each corner of containment
- Emergency Vent on tank
Our storage tank also has the following MITIGATING SAFEGUARDS:
- LEL Detectors and Alarms (both audible and visual that sound both at the tank and in the control room)
- Spring-Loaded FAIL SAFE CLOSED fire valves on all flammable liquid tanks in the secondary containment
- On site fire brigade with extra foam capabilities
In our PHA we ask what can cause us to overflow the tank and what would the consequences of an overflow event entail (it could be very bad – Buncefield Incident). We then consider the risk of this event based on the FREQUENCY and SEVERITY, while taking into consideration our SAFEGUARDS that are in place that would PREVENT, PROTECT, and MITIGATE the incident and it’s consequences. Now here is the trick… IF WE MAKE A CHANGE to any of the items above and this “change” would change our risk profile to the BAD, this would be a change that would REQUIRE US to revisit the node of our PHA where the risk could be impacted.
This is about the only one I could come close to saying it is near 100% a PHA update, but there are many who would argue my position. Anyone else know of a situation that would require an update to a PHA before the five-year revalidation date?