TM54 vs Part L: predicted energy or regulated compliance?
Part L models only regulated energy to prove regulatory compliance; TM54 predicts total in-use energy including everything the occupier will actually pay for.
A Part L compliance model deliberately excludes unregulated loads — small power, servers, catering, lifts, external lighting — and assumes standard operating hours. That is appropriate for a pass or fail test, but it means the compliance figure can be less than half of what the building really consumes. CIBSE TM54 exists to close that performance gap by predicting operational energy with realistic occupancy, equipment and hours.
TM54 vs Part L, attribute by attribute
| Attribute | TM54 | Part L |
|---|---|---|
| Purpose | Predict real operational energy | Demonstrate regulatory compliance |
| Loads included | Regulated and unregulated | Regulated only |
| Operating assumptions | Project-specific, realistic | Standardised |
| Output | Predicted kWh/m²/yr with uncertainty range | BER against TER, pass or fail |
| Audience | Investors, occupiers, net zero reporting | Building control |
| Mandatory | No — increasingly required by clients and frameworks | Yes |
Which one applies to you
Choose TM54 when
- /You are targeting NABERS UK, net zero or a design-for-performance commitment
- /An occupier or investor wants a credible energy budget
- /You need to size plant and renewables against real demand
Choose Part L when
- /Any new build or major refurbishment needing building control sign-off
- /Design-stage and as-built submissions
Running TM54 alongside the Part L model costs comparatively little because the geometry already exists, and it is the only way to know whether a compliant building will also be an efficient one.
Common questions
+Can TM54 be used for Part L compliance?
No. Compliance must be demonstrated through the approved SBEM or DSM route; TM54 is a complementary prediction, not a substitute.
+How accurate is a TM54 prediction?
Typically within a stated uncertainty band rather than a single number, because it depends on occupancy assumptions — which is precisely why the method requires those assumptions to be published.