SFG20 and maintenance standards, explained.

SFG20 appears in a great many maintenance specifications and a great many tender documents, and it is understood in noticeably fewer of them. It is common to find a contract requiring maintenance to be carried out in accordance with SFG20 where neither party holds a subscription and nobody involved can say what the relevant schedule contains. This guide is an attempt to be plain about what it is, what it is not, and how to decide whether you need it.

What SFG20 actually is

SFG20 is a library of maintenance task schedules published by the Building Engineering Services Association. It is widely treated as the industry standard specification for planned maintenance in UK building services. For a given asset type, a schedule sets out the tasks that should be carried out, at what frequency, and with what skill level — so rather than deciding from scratch what maintaining an air handling unit involves, you start from a published position that a large part of the industry recognises.

It is a subscription product rather than a public standard. The schedules are licensed, they are maintained and updated as guidance changes, and access is through the publisher. This matters practically: you cannot legitimately lift SFG20 content into your own system, and a vendor who tells you their software includes SFG20 schedules is describing a licensing arrangement you should ask about specifically.

The classification is the part that earns its keep

The feature most useful to a facilities manager is not the task lists themselves — it is that tasks are classified by criticality, distinguishing work that is statutory from work that is mandatory under other obligations and work that is discretionary good practice. That classification is what lets you answer the question every constrained estate eventually asks: if we cannot do all of this, what is the part we absolutely cannot drop?

You can reach the same place without SFG20. It takes considerably longer, it requires someone who knows the duties well, and the reasoning lives in your own heads rather than in a published source you can point at. That last difference is worth more than it looks during a tender or an audit.

What it does not do

• –It does not tell you what assets you have — that is your register, and SFG20 is useless without it
• –It does not record that the work was done; it is a specification, not a system of record
• –It does not remove the need for a competent person to determine what applies to your building
• –It does not cover every asset type you will own, particularly specialist or legacy plant
• –It is not a legal safe harbour — following it is evidence of a considered regime, not a defence in itself

The second point is the one that causes the most confusion in software conversations. SFG20 tells you what the task is. Something else has to hold the schedule, assign it, record the attendance and keep the evidence — which is the job of your planned maintenance system, and is a separate purchase from the schedules themselves.

What it costs, in practice

Pricing is set by the publisher and is tiered, so the only accurate figure is the one they quote you for your estate. What is worth budgeting for beyond the subscription is the part nobody quotes: mapping your asset register onto the schedules. Every asset has to be matched to a schedule, every schedule has to be checked against what that specific asset actually needs, and the frequencies have to be reconciled with what your team can realistically deliver.

On a mid-sized estate that is a project measured in weeks, not an afternoon. It is time well spent, but it is the reason subscriptions sometimes sit unused: the licence was bought, the mapping was never finished, and the regime carried on running on the old spreadsheet.

Deciding whether you need it

There is a reasonable case both ways, and it turns mostly on the size of your estate and whether you are contractually exposed to the question.

• –Worth subscribing if you tender for work where SFG20 compliance is specified — the cost of not holding it is losing bids
• –Worth subscribing if your estate is large or varied enough that determining task content in-house is a genuine burden
• –Worth subscribing if you need to show a client or an insurer that your regime follows a recognised industry standard
• –Probably not worth it for a small estate of common plant where manufacturer guidance and statutory frequencies cover the ground
• –Definitely not worth it if you will not fund the mapping work — an unused subscription is worse than none, because it creates a belief you are covered

Where FacilityOptix sits on this

Plainly: FacilityOptix does not license or embed SFG20 task content, and does not resell it. If you hold a subscription, you structure your regime here around it — the schedules tell you what the tasks and frequencies are, and this holds them, runs them, and keeps the evidence that they ran. If you do not hold one, you build schedules from manufacturer guidance and statutory frequencies, which is what most smaller estates do.

Either way, the thing SFG20 cannot do is the part that gets estates into difficulty — proving the regime actually ran. That is a question about statutory compliance records rather than about specification, and a perfectly specified regime with no completion evidence fails an audit exactly as badly as an unspecified one.