A safe system of work is the formal procedure that results from examining a task, identifying its hazards and defining a method that eliminates or minimises them. It is an old phrase and a precise one, and it is routinely used as a synonym for having thought about safety, which it is not. A system has parts: a defined method, the people competent to follow it, the equipment it assumes, the checks that confirm it is in place, and the authority to stop when it is not.
How one is built, in order
Start from the task as it is actually performed rather than as it is described. Identify what could go wrong at each step and who would be harmed. Decide the method that removes or minimises each, preferring elimination and engineering controls over instruction and protective equipment. Write the method as a sequence somebody can follow. Then decide how you will know it is being followed, and who can stop the work. The last two steps are the ones that turn a document into a system.
Permits are safe systems with a time limit
A permit to work is not a different animal; it is a safe system of work for a high-risk task, made specific to one occasion and authorised in writing before it starts. Confined space entry, hot work and work on live systems are the usual cases. The permit adds two things to the standard method: confirmation at the time that the preconditions are actually true, and a formal hand-back so the system ends deliberately rather than by everyone drifting away.
Safe work practices are what the system looks like from the floor
The system as written is a manager's artefact; what people actually do is the practice. The gap between the two is where incidents live, and it is measured by watching the work rather than by re-reading the document. Firms that close it well do something unglamorous: they check a handful of tasks regularly, record what they saw, and revise either the method or the training when the practice has drifted for a reason.
Stop-work authority is the part most systems are missing
A system with no explicit authority to stop is a system that assumes nothing will ever be wrong. Naming who can stop the work, making clear that nobody is penalised for using it, and recording when it was used are what make the rest credible. It is also the cheapest control in the whole set, and the one most often absent from otherwise careful documents.
Questions people ask about safe system of work
What is a safe system of work?
A formal procedure resulting from a systematic examination of a task, which identifies the hazards and defines methods that eliminate them or minimise the risk. In practice it is the written method plus the competence, equipment, checks and stop authority that make the method happen.
What is the difference between a safe system of work and a method statement?
Largely context. A method statement is the written description of how a specific job will be done, typically supplied to a client; a safe system of work is the broader term for the procedure the organisation operates, of which the method statement may be the written form. Where a client asks for one, ask which they mean.
What are safe work practices?
The behaviours the system expects on the floor: how the task is actually carried out day to day. They are the test of whether the system exists, because a written method nobody follows describes an organisation that has documented safety rather than achieved it.