Optimising EPC Ratings for MEES in UK Offices 

If you own or manage commercial offices in England & Wales, MEES (Minimum Energy Efficiency Standards) makes the EPC rating more than a box-tick, it governs whether you can let space. Policy is still evolving, but the direction of travel is tighter minimums (currently E, with government signalling a likely uplift to B sometime after 2030 but before 2035, and an interim C around 2028).  

Below is a concise, results-first method to lift your EPC quickly and cost-effectively, starting with data hygiene and moving through end-use targeting. 

1) Get the baseline right (many EPCs are wrong) 

Before buying kit, fix the inputs. Inaccurate geometry, “assumed” system efficiencies, or missing evidence can sink a rating. 

• EPC error rates are non-trivial. A SPEC white paper (2019) reviewing lodged EPCs found ~1 in 4 certificates mis-measured by ≥10% in floor area, enough to swing the grade; Which? has also highlighted systemic inaccuracies (their small sample found widespread errors). Treat any old EPC (especially pre-2022) as suspect and validate.  
• Non-domestic EPCs use SBEM/DSM and default to conservative values when data is missing. If you don’t provide verified manufacturer data (boilers, chillers, AHUs, lighting photometry, SFPs, heat-recovery efficiency, controls), the software will downgrade the model with low default efficiencies. 
• Lighting is commonly mis-scored: if an assessor can’t see lamp types or drivers, fittings default to “inefficient,” even when they’re LED. Ensure access and spec sheets.  
• Tip: Re-lodge EPCs granted before June 2022 if you’ve electrified systems or improved specs. The SBEM v6.1 methodology change and updated carbon factors can materially improve the rating when modelled correctly.  

2) Target the biggest EPC “end-uses” first 

Non-domestic EPCs are carbon-based asset ratings. SBEM breaks demand into heating, cooling, lighting, ventilation/auxiliaries, and hot water—optimise the largest bars in your model first; that’s where the cheapest rating points live. 

How to build your cost-efficient strategy 

1. Pull the SBEM end-use breakdown (your assessor can export it). 
2. Rank end-uses by contribution to EPC emissions. 
3. Map one or two measures per top end-use with verifiable specs (so the asset model credits them). 
4. Iterate the model to test marginal EPC uplift vs. cost. 

3) High-impact, often low-cost measures (that SBEM actually credits) 

Lighting (frequently the cheapest tonnes/£) 

• Swap to high-efficacy LED and evidence it: provide luminaire datasheets so the model uses the real lumens/W, not defaults. Pair with occupancy and daylight controls—these reduce the EPC lighting load directly. 
• Daylight improvements (where appropriate) can further reduce artificial lighting use in the model.  

Cooling & solar gains 

• Solar-control window film can cut solar heat gain and cooling loads without full glazing replacement—often a fast-payback retrofit. Ensure film SHGC data is captured so SBEM can apply it via glazing parameters. 
• Optimise chiller SEER (or VRF seasonal performance) and document part-load curves; model variable-speed fans/pumps to reduce auxiliary energy.  

Heating & hot water 

• High-efficiency boilers (if on gas) with documented seasonal efficiency; or heat pumps where suitable—remember the non-domestic EPC is carbon-weighted, so lower electricity carbon factors post-2022 can favour efficient electrification in the model.  
• Heat recovery on ventilation (documented efficiency) typically scores well in SBEM.  

On-site generation 

• PV reduces grid electricity in the model; make sure inverter and array data are included. 

4) Measures that won’t move the EPC (but help operations) 

EPCs are asset, not operational, ratings. Purely behavioural tweaks (fine-tuning setpoints, staff engagement) won’t shift the EPC unless they are represented as permanent control functions in SBEM (e.g., baked-in schedules/controls). Keep doing them for cost and carbon, but don’t expect a letter-grade jump without modelled asset changes.  

5) A simple, repeatable upgrade plan 

1. Desktop re-model of the current asset with fully evidenced inputs (see checklist) to establish a trustworthy baseline. Given the error landscape, assume you can “win back” points just by correcting defaults and geometry.  
2. Scenario testing: run 3–5 packages targeting the largest end-uses—e.g., 
3. Package A (Quick wins): LED + controls + BMS scheduling + selective window films on solar-exposed façades. (Most buildings can achieve B ratings with only inexpensive, fast payback upgrades  
4. Package B (More intrusive works): Demand reduction, optimising plant efficiency and introducing upgrades such as heat recovery on ventilation 
5. Package C (Electrification): A+ heat pump for heating/DHW + PV. 
   Compare £/EPC-point uplift and capex/lead-times. 
6. Sequence with lease events & refurb cycles to minimise downtime and capture MEES risk early, noting government’s likely uplift towards B next decade.  
7. Re-lodge EPC post-works with all evidence attached. 

6) Policy watch (why you should act now) 

• Government communications and major advisors indicate tighter commercial MEES—likely EPC B—after 2030 but before 2035; an interim C has been discussed around 2028, though details may still shift. Don’t bank on long grace periods. 

Key takeaways 

• Accuracy first: many EPCs are flawed; correcting data and defaults is often the cheapest uplift.  
• Chase the biggest bars: target lighting, cooling/solar gains, and auxiliaries using SBEM end-use outputs.  
• Model what you install: provide hard evidence so SBEM credits your upgrades (LED specs, control narratives, HVAC seasonal data, film SHGC, PV).  
• Get ahead of MEES: expect a ratchet to EPC B next decade; reaching C by around 2028 is a prudent interim goal.