Scottish Energy Performance Certificates - new metrics: research

Potential new metrics

Following initial discussions with SG about the policy aims driving their desire to use a new metric, a number of potentially suitable metrics were proposed, and their pros and cons considered. The key requirement was for the new metrics to be a more direct measure of the building’s energy requirement, regardless of what heat/power sources are used.

Fabric Energy Efficiency

Fabric Energy Efficiency (FEE) is a metric already specified in the SAP methodology document, although it is currently only used in the context of new build assessments (for building regulations); it is not generated by RdSAP software or referred to on EPCs.

For the calculation of FEE, the actual fabric data (U-values etc.) entered by the assessor is used, but instead of the entered space heating and hot water generation system, a standardised direct electric space and water system is assumed in all cases. This ensures that the FEE changes only in response to the fabric specification, allowing a like for like comparison of fabric performance between dwellings, whilst still giving a relatable overall energy consumption total. A benefit of this option is that it is an existing metric which is focussed on fabric performance.

Possible limitations to consider are that FEE doesn’t take into account some features that are usually considered to be a key part of dwelling’s energy efficiency, such as cylinder insulation and the inclusion of how water saving technologies like waste water heat recovery (WWHR) and solar thermal, which are likely to be relevant whatever water heating system is used. If these are to be covered by the separate policy levers relating to heating systems used, that may be acceptable, but if such technologies were to fall between the cracks it is possible this could disincentivise their use. In particular, WWHR systems, which effectively reduce hot water demand, once installed could be seen as an inherent element of the building’s efficiency.

FEE based on local weather

In preliminary discussions, SG suggested it might be desirable to use local weather data when calculating the new metric (in contrast the current EPC rating for which standardised UK average weather data is always used). This would require a minor change to SAP software but is technically easy to achieve. The question is around what impact that would have on ratings and whether it would create any practical difficulties, noting that up to this point local weather data has not been used for any ratings purpose. Existing SAP software does not allow the use of alternative weather data for the ratings or FEE calc, but it does use local weather data when calculating the ‘current running costs’ presented on EPCs, giving a means to test the likely impact of allowing this. Some examples were run for the same poorly insulated example dwelling (an 89m² 3-bedroom semi-detached dwelling with uninsulated cavity walls, and 50mm loft insulation) in each of the 9 regions in Scotland:

This isn’t exactly analogous to how the FEE would vary because it doesn’t include water heating, but it suggests the FEE (if based on local weather) would vary by no more than about ±8% across Scotland.

The implications of having FEE vary with location are that a higher level of fabric insulation would be required in cooler regions to obtain the same FEE rating – i.e. it would be harder to get a C-rating in colder regions. There is some logic for homes in cold places to be better insulated, but it is possible this could be seen as penalising them if in practice this increases the cost of the measures needed to get to a required standard. It also would make it more difficult to offer universal advice on ‘what to do to get you home up to band C’, or similar; so, the acceptability and practicality of this option should be considered carefully.

Of course, it would be possible to use Scottish average weather data, instead of UK average, without this variation. The only potential downside of this would be to reduce direct comparability with ratings in other UK administrations.

The FEE has never previously been recorded in EPC registry entries, so it would not be possible to update old EPCs to use the new metric^[5] – only new EPCs would be able to display the new rating.

Heat loss parameter

The heat loss parameter (HLP) is an intermediate result from the SAP calculation which describes the specific heat loss per unit of the dwelling’s external surface area, in W/m²K. It includes ventilation heat losses as well as conduction heat losses. It is in that respect a very pure measure of the thermal efficiency of the shell of the building. This is therefore another option for use as a fabric efficiency rating metric.

It is perhaps more abstract and less tangible than the FEE (an annual energy consumption figure) and shares the same shortcomings as the FEE in terms of not giving credit for hot water related efficiency features that might be considered relevant. Despite being a long-standing SAP intermediate calculation parameter, it is also not recorded in existing EPC registry data files, meaning it could only be readily used for new assessments.

HLP is not affected by external temperature in the way FEE is, although it does vary with windspeed, which affects ventilation losses. Using an example, the HLP in Shetland (the windiest region) was found to be 8% higher than when the same dwelling was located in the (least windy) Borders region, suggesting a range of about ±4% across Scotland.

Heat loss coefficient

The heat loss coefficient (HLC) is another existing parameter in the SAP calculation that directly represents fabric heat losses. It is similar to the HLP but is not normalised for the surface area of the dwelling, so would result in larger homes receiving worse ratings than smaller ones for the same level of fabric energy efficiency. If this was to be used as a metric it would therefore essentially mean large homes would need to be more efficient to get the same rating as smaller ones, which is likely to be difficult to justify. Following initial discussions, this option was not considered further.

Modified FEE or HLP calculations

It could be possible to modify the FEE or HLP in some way to form a new metric which takes into account other features, e.g., relating to hot water efficiency. FEE is a measure of energy consumption, so the energy impact of such features could be calculated to create an adjustment – e.g. the hot water saving from WWHR could be calculated and subtracted from the total when FEE is calculated. Alternatively, the FEE calculation could perhaps be modified so it uses some features of the actual system (e.g. cylinder insulation), while others are standardised (hot water source). The difficulty with this is that some features of interest are possible in combination with one type of system, but not another. For example, for cylinder insulation to be recognised there would have to be a cylinder present (which is not the case with combi boilers). If different system types are assumed for different situations, the consistency of approach which allows like for like comparisons is quickly lost; so, this would probably be quite a complicated and possibly confusing approach.

It is hard to see how any adjustments relating to hot water systems could work if HLP is used, since this metric has units that are very specific to fabric losses (W/m2K), so this option was not considered further.

Useful energy consumption

The sum of the heat required for space heating and hot water (aka ‘useful energy’, or ‘heat demand’), before heat generation efficiency is applied, per square meter of floor area, would give another possible measure of the efficiency of the building. This could take into account all features other than the heating system efficiency, including hot water system saving features (if desired), so might be a good option if these are seen as important. The disadvantages are that things like whether a combi or regular boiler is used, and the type of heating controls present also impact useful energy, so it is not possible to disentangle some features that are specific to the heating system and might no longer apply if a new heating system was installed.

Options for what energy uses could be taken into consideration in a useful energy metric are discussed further below.

Delivered energy consumption

Summing the fuel requirements (i.e. the energy use after the application of the heating system’s efficiency) to give the overall fuel requirements of the home for the regulated uses included in SAP would give a wholistic assessment of the energy efficiency of the dwelling and its energy supply systems. However, this is not aligned with the key policy intention to separate fabric efficiency from heating system efficiency. Another potential issue is that homes with heat pumps could get a very favourable rating by this measure on account of their coefficients of performance being much great than 1, even with relatively poor fabric efficiency. However, this option could still be considered further if SG later find that it is going to be problematic to keep to two aspects separate and there is a change of policy direction.

What items could a useful energy metric include?

Following discussion of the above options with SG, the use of a metric from SAP which expresses the energy needs of the dwelling prior to generation efficiency being applied was favoured as a measure of the building’s energy efficiency. In SAP terms this is usually referred to as the ‘useful energy’ (UE) requirement of the building. However, this is a general term and has some optional components, which may or may not be appropriate for the proposed purpose.

Potentially relevant intermediate results from a SAP 10.2 calculation are as follows, where the number in brackets represents the SAP 10.2 worksheet number:

1) Output from water heater^[6] in kWh/yr (64)

2) Electricity used for instantaneous electric showers (64a)

3) Space heating energy required in kWh/yr (98a)

4) Space cooling energy required in kWh/yr (108)

5) Electricity used for pumps and fans (231)

6) Electricity used for lighting (232)

7) Electricity saved (used within the dwelling) from PV (233a)

8) Electricity saved (used within the dwelling) from wind turbines (234a)

9) Electricity saved (used within the dwelling) from hydro (235a)

10) Electricity saved (used within the dwelling) from micro CHP (235c)

11) Electricity generated and exported from PV (233b)

12) Electricity generated and exported from wind turbines (234b)

13) Electricity generated and exported from hydro (235b)

14) Electricity generated and exported from micro CHP (235d)

15) Energy saved or generated by Appendix Q items (236a, b, c…)

16) Energy used by Appendix Q items (237a, b, c…)

Items 5) and 6) shaded green depend on the heating system used, so after some initial consideration it was decided to exclude these.

Items 7) to 14) relate to on-site generation. Discussions with SG concluded that this would be inappropriate to include, e.g. to avoid the situation where a dwelling with inefficient fabric could get a high rating by having lots of on-site generation.

It is possible Appendix Q items (15 and 16), i.e. new technologies not yet in SAP, could also be a new form of on-site generation, or otherwise relate to heating system efficiency and therefore be unsuitable for this rating; so there is an argument for excluding these too.

On the other hand, it would be a shame to exclude valid new fabric technologies from the rating. The difficulty here is that, by definition, we don’t know what these will be in advance.

To give a feel, the current technologies recognised through Appendix Q are:

• Smart hot water cylinder (increases self-use of PV and minimises standing losses)
• Smart air-brick (improves floor U-value and air tightness)
• Solar pre-heater of ventilation system air
• Solar assisted heat pumps

Three of the four listed technologies relate to renewable generation, so on balance it is probably better to exclude appendix Q items to avoid these, or future technologies, impacting the rating in an undesirable or unforeseeable way. So the proposed approach is only to include items from the following four useful energy components:

1) Output from water heater in kWh/yr (64)

2) Electricity used for instantaneous electric showers (64a)

3) Space heating energy required in kWh/yr (98a)

4) Space cooling energy required in kWh/yr (108)

The sum of the energy use of some or all of these items would therefore give the measure of the performance of the building. Limiting it only to space heating and cooling energy would result in a wholly ’fabric only’ measure, whereas including water heating and the other sources would extend this to potentially encourage other energy efficiency options, if desired.

The resulting energy consumption figure should be normalised for floor area in some way (prior to being mapped onto an A-G scale) to prevent large homes automatically getting poorer ratings than small ones. The simplest way of doing this is to divide the total useful energy figure by the floor area.