RAMS: Risk Assessment and Method Statement Explained

Before anything else, a disambiguation, because "RAMS" is one of the more unfortunate acronym collisions in engineering. In construction, contracting and facilities management, RAMS means Risk Assessment and Method Statement: the safety paperwork a contractor submits before being allowed to start work. In reliability engineering, systems engineering and rail, RAMS means Reliability, Availability, Maintainability and Safety: a set of engineering attributes designed into a system and demonstrated over its life. The two have almost nothing in common beyond four letters. This article is about the first one. If you came looking for the reliability sense, start with reliability versus availability versus maintainability, the reliability engineering guide, or the metrics behind them. The dependability vocabulary that underpins that sense of the term sits in IEC 60050-192:2015, the dependability part of the International Electrotechnical Vocabulary, which is a voluntary international reference rather than law anywhere.

The message up front: nothing about RAMS is a distinct document type. It is two documents, produced for different purposes, submitted and read as one package. The single most informative property of a RAMS submission is whether the two halves agree with each other, and the single biggest cause of worthless RAMS is a client process that demands them faster than anybody could honestly write one.

This is a general explanation of how RAMS work as a process. What any specific contract or site actually requires is set by that contract, by the site's own safety management system, and by the law of the jurisdiction the work is in. Check those before you rely on anything here.

1. What RAMS actually is

RAMS is a packaging convention. Before starting work a contractor is asked to demonstrate two things: that they understand what could go wrong, and how they intend to do the job so it does not. The first is a risk assessment, the second a method statement. Bundled, submitted and reviewed together, they are called RAMS.

The contents of each half are owned elsewhere. A risk assessment identifies hazards, evaluates the risk arising from them and records the controls that will reduce it; the process is covered in the risk assessment guide. A method statement describes how the work will be carried out, in sequence, with the resources, equipment and people it needs; see the method statement guide. If the difference between them is what you are pinning down, the risk assessment versus method statement comparison does exactly that job. What follows is the part nobody else covers: the pair as a deliverable, and the relationship between the two parties who exchange it.

It is worth being honest about the term's geography. RAMS is overwhelmingly a British and Commonwealth industry expression, strongest in UK construction, building services contracting and facilities management, and carried outward by British and Irish contractors and FM operators into Ireland, the Gulf, parts of Africa, Australia and New Zealand. In much of the United States the same package travels under other names: a job hazard analysis or job safety analysis with a work plan or safe work procedure attached. A North American safety manager may well have never heard the acronym while running an equivalent process daily, so in a mixed-nationality project team, say "the risk assessment and the method statement" once and you will have saved a meeting.

2. No standard or regulation defines "RAMS"

This needs stating plainly, because the acronym's ubiquity gives it a false air of statutory status. No standard and no regulation in any jurisdiction I am aware of defines RAMS as a deliverable. No clause anywhere says a document called RAMS must exist, must contain particular sections, or must be approved by anybody. The term is industry custom that hardened into procurement language.

What legislation names instead is the underlying duty. In Great Britain the duty to assess risk sits in the Management of Health and Safety at Work Regulations 1999 (SI 1999/3242), Regulation 3, under the framework of the Health and Safety at Work etc. Act 1974 (c. 37); for construction work in Great Britain the planning and coordination duties sit in the Construction (Design and Management) Regulations 2015 (SI 2015/51). None of those instruments mentions RAMS. In the United States, federal 29 CFR 1910.132(d) requires a workplace hazard assessment with a written certification for PPE selection in general industry, about as close as US federal law comes to a mandated written assessment, and again says nothing about method statements. In the United Arab Emirates, occupational safety duties for private sector employers sit in Federal Decree-Law No. 33 of 2021 on the regulation of employment relationships, Article 13, administered by MOHRE, with emirate frameworks layered on top; in Abu Dhabi that framework is the Abu Dhabi Occupational Safety and Health System Framework (ADOSH-SF), Version 4.0, administered by the Abu Dhabi Public Health Centre. Where a Gulf client demands RAMS, that demand is contractual and derived from British practice rather than from the local instrument.

On the voluntary side, ISO 45001:2018 (as amended by Amd 1:2024) sets certifiable requirements for an occupational health and safety management system, but does not name RAMS either. Its clause 8.1.2 requires the hierarchy of controls, the principle your risk assessment should be applying, also required by ANSI/ASSP Z10.0-2019, section 8.4 in the United States and described on a free NIOSH page; see the hierarchy of controls guide. ISO 31000:2018 gives risk management guidelines, and it is guidance only: no auditable requirements, and no accredited ISO 31000 certification exists for an organisation, whatever a training provider tells you. IEC 31010:2019 catalogues risk assessment techniques; note the designation is IEC 31010, not ISO 31010. All of these are voluntary and law nowhere by themselves.

3. Why the pair travels together

The reason the two documents are bundled is not administrative laziness. Either one alone leaves the obvious next question unanswered, and a client deciding whether to let a stranger work on their premises needs both answers.

A risk assessment alone tells the client the contractor can name the hazards: working at height, live electrical parts, hot work, confined space, a shared occupied corridor. It does not tell them whether the job as the contractor intends to perform it is compatible with those controls, or with this building. Naming a hazard is not the same as having a plan.

A method statement alone tells them the intended sequence: isolate, verify dead, remove the panel, replace the component, reinstate, test, hand back. It does not tell them whether anyone thought about what can go wrong at each step, or whether the sequence was written by somebody who understood the hazards or somebody copying a previous job. A method with no assessment behind it is a plan with no reasoning attached.

Together they answer both halves: what can hurt someone here, and what we intend to do about it while getting the work done. That is the entire logic of the pair, and why clients ask for both even when their procedure names only one.

4. Consistency across the pair is the most informative thing in a submission

This is the substantive point of the whole article, and it is the one thing I would want a reviewer to take away. The two documents must be consistent with each other, and their consistency tells you more about the quality of the submission than any individual section does.

Think about what it means when they disagree. If the risk assessment identifies a fall hazard and specifies a mobile elevating work platform as the control, and the method statement describes standing on a stepladder, then either the assessment was written without reference to the method or the method was written without reference to the assessment. In either case the two were produced separately to satisfy a requirement rather than produced together to plan a job. That is reliable evidence, not a formatting nitpick. The reverse case is equally telling: if the method describes equipment, a space or an adjacent system that appears nowhere in the assessment, the assessment did not consider the work as it is actually going to be done.

The value of a consistency check is that it is cheap, fast and cannot be gamed by adding pages. A reviewer with fifteen minutes and this table will learn more than a reviewer with two hours and a section-presence checklist.

5. The RAMS lifecycle, stage by stage

RAMS is better understood as a process between two parties than as a file. Here is the lifecycle, and where each stage typically breaks.

Request and scoping. This is where client-side failure begins, and since much of the audience for this topic sits on the client side it is worth dwelling on. A contractor cannot write a site-specific document about a site they have not seen and have been told nothing about. If the request is "send RAMS" against a job description of "repair AHU", the contractor has no choice but to send something generic. A useful request carries the actual scope, the location and access arrangements, the building's own rules, what permits and isolations will be required, who else is working nearby, the occupancy and operating constraints, and an offer of a site visit. Clients who provide nothing and then complain that submissions are generic have built the outcome they are complaining about.

Production. The pair should be shaped by somebody who knows how the work is actually done: a supervisor who has performed the task, not an administrator assembling clauses. Involving the crew who will do the job is the cheapest quality improvement available, because they will immediately name the step the office version glossed over. The task-level analysis techniques in the job safety analysis guide are a good way to structure that conversation.

Submission, and the lead time problem. This deserves naming honestly. RAMS demanded at the last minute get copied. There is no version of events in which a contractor asked on a Thursday afternoon for documents covering a Friday morning job produces a considered, site-specific pair; they will find the nearest previous submission and change the header, because that is the only available response. So a client whose procedure requires RAMS days before work, but whose planning routinely raises jobs at short notice, has built a system that guarantees generic documents while formally prohibiting them. The contradiction sits in the client's process, not the contractor's ethics. If you want considered submissions, job release has to happen far enough ahead that considering is possible, and where it cannot, you need the pre-agreed arrangements discussed below rather than a fiction of bespoke paperwork.

Review and approval. Covered in its own section below, because the gap between competent review and compliance review is where most of the value is lost.

Acceptance and briefing. Approval means nothing until the people doing the work know what is in the document, and the people who turn up are frequently not the people who wrote it: subcontracted labour, agency operatives, a substituted crew after somebody called in sick. Briefing is the moment the document stops being paperwork and becomes a shared plan, and it is the moment most commonly skipped or reduced to a signature sheet. A toolbox talk at the point of work, referencing this specific method rather than the hazard in general, is the practical mechanism. If the crew cannot tell you the sequence and the two or three things that would stop the job, they have not been briefed.

Use and deviation. Conditions on the day will differ from those assumed: a valve is not where the drawing says, the isolation cannot be proved, the access route is blocked, the fixing is corroded solid. What matters is that a route exists to change the method with authority, and that the route is faster and easier than departing from it quietly. If changing the method means a two-day resubmission and a call to someone who will be annoyed, the crew will improvise and say nothing, and your approved document now describes a job that did not happen. Stop-and-consult behaviour is a design property of the process, not a character trait of the operatives.

Close-out and retention. When the work finishes, the pair becomes a record: what was planned, what was briefed, what changed and who authorised the change. That record has three genuine uses. It supports investigation if something went wrong. It is the starting point for the next occurrence of the same task, which is how a generic document earns the right to be generic. And it evidences the arrangements you had in place, which is what an insurer, an auditor or an enforcing authority will ask about. Retained without any of those uses in mind, it is just storage.

6. Competent review versus compliance review

Two entirely different activities share the word "review", and most organisations do the second while believing they do the first.

A competent review reads the documents against the site and the job. Does the described method work in this building, with these access constraints, around these occupants? Do the controls actually address the hazards identified, and are they at a sensible point in the hierarchy rather than defaulting to PPE and instruction for a problem that could be engineered out? Are the interfaces with the client's own systems addressed, particularly permits, isolations, hot work controls, roof access and shutdown arrangements? Is competence evidenced rather than asserted? Are the emergency arrangements real for this location, naming a rescue plan that works here rather than a line about calling the emergency services? A competent review produces comments, questions and sometimes conditions.

A compliance review reads the documents against a template. Are the required sections present, is it signed, is it in date, is the insurance certificate attached, is the risk score below the escalation threshold? A compliance review produces a tick.

The honest observation is that most RAMS review is the second kind, and the effect is predictable. Contractors are commercially rational: they learn what gets approved and produce more of it. If length and section coverage get approved, submissions get longer. If a low residual risk score gets approved, scores get lower. The system trains its suppliers to produce fatter documents rather than better methods, and then everyone complains that RAMS have become a paper exercise. They became a paper exercise because they were reviewed as paper.

7. The client-side accumulation problem

In facilities management this is the reality, and it deserves its own treatment. A site of any size holds hundreds of contractor RAMS: lift maintenance, HVAC, fire systems, generators, water treatment, cleaning, pest control, facade access, electrical testing, minor works, plus every one-off project. Most were reviewed superficially because no team has the capacity to review that volume properly. Many describe equipment, personnel or site arrangements that have since changed. A good number were approved by someone who has left. Taken as a whole, the collection functions less as a safety control than as a liability archive: a large body of documents the organisation formally endorsed and cannot now stand behind in detail.

The instinct when this is pointed out is to review everything, which fails immediately because the capacity does not exist and never did. The better answer is proportionality, and it is the most practically valuable idea here.

• Risk-proportionate review depth. Accept openly that not every job justifies the same scrutiny. High-hazard work, confined space entry, live electrical work, facade access, hot work, lifting operations, anything touching life safety systems, gets a genuinely competent review by someone who knows that work. Low-hazard routine attendance gets a light check and a standing arrangement. Writing that tiering down, agreed by whoever owns the risk, converts an impossible uniform standard into an achievable differentiated one.
• A register with status, not a folder with files. You need to answer quickly which contractors have current accepted arrangements, for what scope, reviewed by whom, with what conditions attached. A folder of PDFs cannot answer that; a simple register can, and it is usually the single highest-value administrative change available. What matters is that status is visible, not which product stores it.
• Re-review triggered by change, not by the calendar. An annual re-review of an unchanged arrangement for an unchanged task consumes the capacity you needed for the jobs that did change. Trigger re-review on what actually invalidates the document: scope, equipment or access changes, a change of personnel or subcontractors, a change to the building, an incident or near miss on that task, or a revision to the contractor's own procedure.
• Deliberate scope acceptance. For repeated routine work, accept an arrangement for a scope rather than re-approving a near-identical document each visit, and put the effort into a site-specific confirmation at the point of work instead.

None of this lowers standards. It recognises that review capacity is finite, and that spreading it uniformly guarantees the jobs which needed real scrutiny did not get it. The same proportionality logic runs through contractor management generally, covered from the operations side in the facilities maintenance management guide, and the tiering usually maps onto how you already classify jobs, as set out in the work order types guide.

8. Generic versus site-specific, treated fairly

"Generic RAMS" is used as an insult, and it should not be. For a genuinely repeated task performed the same way in comparable locations, a well-maintained generic risk assessment and method statement, kept current and actually reflecting how the work is done, plus a site-specific confirmation step at the point of work, is a legitimate and often better answer than pretending to write something fresh each time. It is better because the generic document gets attention it would never receive as one of two hundred bespoke submissions a year, and because a short honest site confirmation is more likely to be read than a forty-page pack.

The failure is not genericity. The failure is a generic document presented as site-specific. That is a misrepresentation, and usually easy to spot.

• Plant references or asset tags that do not exist here, or belong to a different site.
• A hazard list covering conditions absent here, showing the assessment was written for a superset of sites.
• Access, parking, permit or induction arrangements that are not yours.
• No mention of a client control that plainly applies, such as a permit regime or a required isolation.
• Emergency arrangements naming no location, assembly point or rescue route.
• Personnel named who are not the attending crew, or a competence list unrelated to the submitted team.
• A revision date and reviewer predating the enquiry, under a header claiming it was prepared for this job.

The constructive response to a generic submission is not rejection with a demand for something bespoke. It is to say: this is a reasonable base document, now tell me in writing the three things that are different about this building and this job. That is answerable in an hour and produces something genuinely useful. A demand for a fresh forty-page pack by tomorrow produces another copy.

9. Reactive and emergency work

Most articles on RAMS quietly assume planned work, which is why they are of limited use in maintenance. A large share of facilities work is reactive: a leak, a failed pump, a stuck lift, a lost supply, a fault that has taken a critical service out. There is no time to write, submit, review and approve RAMS before a crew attends, and pretending otherwise produces either a fictional document backdated to look compliant or a delay that makes a bad situation worse. The realistic answer has three parts.

• Pre-agreed arrangements for foreseeable reactive work. Reactive is not the same as unforeseeable. You know pumps fail, drains block, lifts trap passengers, leaks happen. The categories are predictable even though the occasions are not. Agree assessments and methods for those categories in advance, in slow time, with proper review, so the arrangement already exists when the call comes. This is the single most effective thing a client can do about reactive RAMS.
• A dynamic assessment on arrival. The attending person assesses the actual situation against the pre-agreed arrangement before starting, and records that they did. Kept short and genuinely done, this is not paperwork theatre: the point is a deliberate pause to check the situation is the one the arrangement anticipated.
• A defined stop-and-consult trigger. The crucial element. The arrangement must state plainly what means "this is outside what was agreed, stop and call": conditions the method did not anticipate, hazards outside the scope, work needing a permit nobody has raised, anything requiring a competence the attending person does not hold. The escalation route then has to be genuinely available and quick, because a stop trigger that leads to an unanswered phone teaches people not to stop.

Where reactive work turns out larger or more hazardous than the pre-agreed arrangement covers, it should convert into planned work with proper documents once the situation has been made safe. Making safe and completing are two different jobs, and conflating them is how reactive attendance quietly swallows work that deserved planning.

10. RAMS and permits to work

RAMS and a permit to work are different instruments and neither replaces the other. The pair is planning: produced ahead of the work, describing the intended method and the reasoning behind it. A permit is authorisation at the point of work: a controlled, time-bounded instrument issued by the party responsible for the plant or the area, confirming that specified precautions are in place right now, for this task, in this location, for this period.

The relationship is that the pair informs the permit. A competently reviewed method statement tells the permit issuer what precautions the work requires, what isolations must be proved, what other activities conflict, and what the handback condition should be. Where a client's permit regime and the contractor's method have not been reconciled, you get the familiar site failure: a method assuming an isolation nobody arranged, or a permit issued against precautions the method never intended to apply. Reconciling the two is exactly what a competent review should be doing, and it is one of the checks in the consistency table above. The mechanics of permit systems are covered in the permit to work guide, and the systems side of running one alongside a maintenance platform in the permit to work and CMMS integration article. Worth knowing: no international standard specifies a permit-to-work system, and the widely used reference is UK HSE guidance HSG250 (2005), written for the petroleum, chemical and allied industries; it is guidance for Great Britain and creates no duties of its own, and it is a free download.

For the wider frame that all of this sits inside, the HSE overview covers how assessment, method, permit and briefing fit together as one system rather than four unrelated pieces of paperwork.

11. How RAMS fails

The failure modes are consistent across operations, and few of them are document quality problems in the narrow sense.

• Demanded late, so copied. One of the most common failures, and among the most fixable, because it lives entirely in the client's planning process.
• Reviewed for presence, not quality. Sections counted, signatures confirmed, method never tested against the site.
• The crew never briefed. The document is approved, filed, and unknown to the people holding the tools.
• The two halves contradict each other. Proof they were produced separately, and usually a sign nobody read both.
• The job changes after approval. Scope grows, access alters, a different component is involved, and the approved documents quietly stop describing the work.
• The same pair reused for a different building. Genuinely dangerous, because the hazards that differ between buildings, access, adjacency, occupancy, existing systems, are exactly the ones a reused document is blind to.
• Approval treated as a formality by both sides. The contractor submits to clear a gate, the client stamps to clear a gate, and the exchange transmits no information in either direction.
• Volume optimised because volume is measurable. The honest organisational failure. Percentage of jobs with RAMS on file, average turnaround, number approved this month: easy to count and easy to report. Whether the method was sound, whether the crew understood it, whether it matched the site: none of these aggregate into a number. Systems optimise for what they measure, so they optimise for volume and then express surprise that quality did not follow. If you want quality you have to inspect it directly, in samples, on site, and accept that your headline metric will never reflect it.

The idea to walk away with

RAMS is not a document, a standard or a legal requirement. It is a conversation between a client and a contractor about a specific job, conducted in writing, in which the contractor explains what could go wrong and how they intend to work, and the client decides whether that is acceptable on their premises. Everything valuable about RAMS lives in the quality of that conversation, and everything worthless comes from treating it as a document exchange.

The practical consequences follow. Ask properly, with real information and real lead time. Check that the two halves agree, first, because it is the fastest quality signal available. Review deeply where the hazard justifies it and lightly where it does not, and write that tiering down so it is a decision rather than an accident of capacity. Accept good generic documents for repeated work with a genuine site-specific confirmation. Plan for reactive work in advance. And brief the crew, every time.

Final thoughts

The most useful shift I can suggest to anyone on the client side is to stop asking whether the RAMS is compliant and start asking whether it is true. Does it describe this site, this job, this crew and this method? Would the people doing the work recognise it as a description of what they are about to do? Those two questions catch nearly everything a section-presence checklist misses, and they take less time. And if your site holds hundreds of approved contractor RAMS nobody could defend in detail, the answer is not a review campaign but a proportionate system: work tiered by hazard, a register that shows status rather than a folder that stores files, re-review driven by change, and pre-agreed arrangements for the reactive work you know is coming. That is less paperwork and considerably more safety, which is usually the sign you are moving in the right direction.

Primary sources worth going to directly rather than reading second hand: the UK Health and Safety Executive for Great Britain guidance including HSG250, ISO for ISO 45001:2018 and ISO 31000:2018, and IEC for IEC 31010:2019 and IEC 60050-192:2015.

Related reading: What is a method statement, Risk assessment guide with examples, Risk assessment vs method statement, Permit to work guide, Job safety analysis, Hierarchy of controls, Toolbox talks, What is HSE, Reliability vs availability vs maintainability (the other RAMS).