If you run an Upper Tier COMAH site, you have to produce a COMAH Safety Report, and the Predictive Aspects section is where your risk assessment has to stand up. The CDOIF MHHHRA guidance changes what that section needs to show. This page explains how MHHHRA affects your Predictive Aspects, the less obvious knock-on to your Technical Aspects, and the practical impacts to expect, so you can see where the work is before an inspector does.
Where the Predictive Aspects sit in your COMAH Safety Report
The Predictive Aspects section is where you demonstrate the risk from your major accident hazards and show it is tolerable. It sits alongside the Descriptive, MAPP and SMS, Technical, Environmental and Emergency Aspects of the report, but it is the section that carries the human harm risk assessment, and it is the one MHHHRA speaks to most directly.
Inside the COMAH Safety Report
Descriptive
Aspects
Predictive
Aspects
your risk assessment
MAPP and SMS
Aspects
Technical
Aspects
Environmental
Aspects
Emergency
Aspects
Every section matters. The Predictive Aspects is the one that carries the human harm risk assessment, and the one MHHHRA speaks to most directly.
What is actually in the Predictive Aspects
1
Major accident scenario identification
Every scenario that could cause a major accident, and being able to show how you arrived at the list.
2
Developing a representative set
You cannot assess everything fully, so a subset is chosen to stand for the rest. How you choose it has to be justified.
3
Consequence modelling
The use of models such as PHAST to understand the extent and severity of fire, explosion and toxic releases.
4
Frequency and likelihood
How often it could happen. Fault trees, event trees and failure rate data, related to your actual assets rather than lifted from a generic dataset.
5
Risk summation
The risk to a person across the whole site, added up, rather than one scenario at a time.
6
Occupied buildings risk assessment (OBRA)
The risk to people inside occupied buildings within the hazard range.
7
ALARP demonstration and further measures
Whether a further measure is reasonably practicable, and if not, why not. Cost-benefit analysis where the judgement is close.
How MHHHRA raises the bar on your Predictive Aspects
MHHHRA is about meeting the HSE’s expectations for the COMAH risk assessment that sits underneath your Predictive Aspects. In practice that means a clearer focus on specific measures rather than generic assessment, and a demonstrable link to relevant good practice. If your Predictive Aspects currently lean on generic numbers or a broad-brush treatment, that is where the guidance will bite first.
The overlooked link: Predictive and Technical Aspects
In my experience across a range of sectors and industries, one of the most common issues the HSE raise with COMAH Safety Reports is a lack of linkage between the Predictive and Technical Aspects: the two are too often written as if they were separate exercises. MHHHRA is not guidance on writing a COMAH Safety Report, but because it forces a focus on specific measures and links to relevant good practice, it naturally pulls those two sections together, which is exactly the gap that tends to get flagged. Handled well, it is a chance to close a long-standing weakness rather than just add work.
A caveat before the comparison. Sites are not all in the same place, and some Predictive Aspects sections are already close to what MHHHRA expects. The left-hand column below is what I most often see across the sites I work on. It is my average, not a claim about yours.
How most reports do it now
What MHHHRA expects
MAH identification and the representative set
A representative set exists, but without full site screening it may not capture everything. Representations are generic, “loss of containment of a flammable from a bulk store”, and made against a substance or an outcome rather than a unit operation. Often derived from a single study: a HAZID, or a HAZOP.
Full site screening shows every potential MAH has been considered, and that the representative set is valid. Representations are specific, “loss of containment of toluene from Tank 101”, made against the unit operation, with a clear indication of what else they stand for. Drawn from HAZID, HAZOP and past accident review.
Impact
More screening work, which may surface scenarios you have not identified. That pulls consequence modelling and frequency assessment along behind it.
Consequence modelling
Extent and severity based largely on fatality or greater.
The COMAH serious danger criteria: fatality plus multiple serious injuries.
Impact
The modelling techniques do not change, and neither do your modellers. But more scenarios means more modelling, and you need a methodology for accounting for injuries, including how many injuries are equivalent to a fatality.
Frequency and likelihood
Because the representation is generic, the frequency inherits that: a generic figure from industry data. The focus lands on the scenario risk number alone.
Because the representation is a specific unit operation, likelihood draws on generic data where appropriate and frequencies derived by analysing the specific causes, through LOPA, fault tree and similar. Sitting inside that analysis is how the unit operation compares to relevant good practice, so the demonstration rests on the number and on compliance with RGP.
Impact
The time and effort of performing the RGP review, and deriving frequencies where that is possible.
The link to your Technical Aspects
Technical Aspects stands alone: an assortment of examples from across the site, aligned to the SRAM. One relief study, one SIL study, one human factors study. You put your best foot forward.
Technical Aspects still carries the general measures, the ones that do not change by scenario. But the Predictive now holds the specific measures inside the scenario. For Tank 101: the relief study showing it is sized for the fire case, the LOPA showing the SIL determination for overfill protection, and the functional safety documentation showing how that SIL was designed and installed.
Impact
In principle, nothing new. This work should already exist. Where it does not, the linkage is what exposes it. And no site can submit every technical document for every unit operation and every scenario, so there has to be a way of showing they are present without sending them all. The guidance does not tell you what that is.
Risk summation
An unlucky person principle. Cumulative individual risk of fatality summed and reported, with societal risk where relevant. Risk fractionated for the time the person is actually at work. Complex sites separate further by location or job role, since not all risks reach all people.
Much the same method, but it has to carry the injury element too. And fractionation is being challenged: sites need to report both on a 24/7 basis and fractionated.
Impact
Not really more work, just a different way of looking at the risk. In theory it pushes existing numbers up. The HSE’s position is that a more accurate analysis, done earlier, should bring them back down.
Occupied buildings risk assessment (OBRA)
Impact
It simply inherits whatever the upstream numbers do.
ALARP and further measures
Largely unchanged, but with a sharper focus on cost-benefit analysis, and on the degree to which the site complies with relevant good practice.
Impact
Little new method. More weight on two things you may not have evidenced heavily before.
Common questions
Click a question to open it.
Does MHHHRA tell me how to write my COMAH Safety Report?+
No. MHHHRA is not safety-report guidance. But it shapes the risk assessment your Predictive Aspects depend on, and it helps close common gaps, including the link to your Technical Aspects.
What is the link between the Predictive and Technical Aspects?+
The Predictive Aspects assess the risk; the Technical Aspects cover the engineering and controls that manage it. They should tell one story, but reports often treat them separately. A tighter focus on specific measures and relevant good practice ties them together.
Will MHHHRA mean more scenarios in my report?+
It can. A fuller screening process can surface scenarios your current report does not cover.
Next step
Predictive Aspects review
An independent read on whether your Predictive Aspects will stand up
I sat on the CDOIF working group behind this guidance, offering review and challenge as it took shape. Before that, 25 years in high-hazard process safety, on operating sites as much as in consultancy, and across a range of sectors. So this is not a reading of the guidance. It is the view from inside the room, applied to your report.
What already meets it
Most of your Predictive Aspects probably does. You should not be reworking what is already sound.
Where the gaps actually are
And what closing them involves, on your timetable rather than an assessor’s.
The link to your Technical Aspects
The gap the HSE most often raises, and the hardest one to see from inside your own report.
You already have a COMAH Safety Report, and people who wrote it. What you may not have is someone from outside it. Fill in the form below with a little about your site and I will come back to you.
