Energy Efficiency Standard for Social Housing: peer review

4 Validation of Existing Retrofit Examples

4.1 Modelling Overview

The consultation document was accompanied by a set of draft retrofit examples, demonstrating how the standard might be met in different property types. Changeworks re-modelled these retrofit examples and validated the inputted energy modelling data presented. Each retrofit example was re-modelled (using both gas and electric heating), and changes noted. The outcomes for re-modelled retrofit examples are provided at Appendix F. Detailed commentary and validation is provided for each example in the Report Annex: Review and validation of Consultation Case Studies (now Retrofit Examples) .

At the time of re-modelling these examples, the new version of RdSAP 2009 (v9.91) was not available (this has subsequently been launched); therefore, RdSAP 2009 (v9.90) was used to validate the examples. This was a much newer version of the software than was used by the Scottish Government ( SAP 2005 (v9.83). Aside from the minor issues set out in table 2 below the baseline assumptions were deemed to be valid by Changeworks' modeller and were therefore not altered ^[16] . However, despite the consistency in inputs it became apparent that that the outputs were sensitive to software changes. It is anticipated that the difference between version RdSAP 2009 9.91 and 9.90 will be less significant than versions RdSAP 2005 v9.83 and RdSAP 2009 v9.90.

4.2 Changes to Modelling

Only minor changes were made to the input assumptions. These related to example 11E where the roof area utilised for PV was boosted from 10% to 20% to accommodate an adequate system output. This was consistent with the other retrofit examples utilising this technology. Some rogue values were in the text for retrofit example, 12G in relation to the 2050 advanced measures. No advanced measures could be identified so values in this column were deleted. Meanwhile, examples 16E, 16G, 17E and 17G saw changes to the baseline assumption on floor insulation, because this would not have been specified by building regulations or likely to have been installed prior to 1990. Table 2 summarises changes identified as part of the validation exercise.

Changes to Case Studies (now Retrofit Examples) in the Consultation Document

Table 2: Changes to Case Studies (now Retrofit Examples) in the Consultation Document

4.3 Alternative approaches and assumptions

Apart from the above changes, the validation exercise identified alternate approaches and assumptions which are discussed in the Report Annex: Review and validation of Consultation Case Studies (now Retrofit Examples) for each example. For example, alternate baseline values could be specified for chimneys and hot water tank insulation but it was found that overall impact wouldn't change the EI or EE rating band of the properties. Alternatives to the retrofit examples were modelled to explore and illustrate sensitivity; these are discussed in more depth in Annex A. Whilst there was some movement in the EI and EE rating it wasn't felt sufficient to warrant remodelling all the examples. Instead social landlords should be aware of this factor, when considering the baseline of their stock.

The depth value specified for loft insulation installed in the period between the baseline and compliance between 1990 and 2015 didn't correspond with building standards in the later part of this period. Instead it matched the SHQS minimum standard of 100mm. Given the lack of data on loft insulation levels and variations due to install date it was felt to be safer to retain this as a worst case scenario. In practice many social landlords and installers will have installed up to 250mm. For these landlords the top-ups specified in the retrofit examples wouldn't be required.

4.4 Overview of impacts of the new version of RdSAP

The following summary table shows where the EE or EI band has changed when re-modelled using a more up-to-date version of RdSAP. Areas in grey indicate no change, values in white show original values which have been altered. For the corresponding changes red values indicate lower bands and green indicates a higher band.

Table 3 Summary of re-modelling changes to SAP and EI bands

A full breakdown and comparison of the changes in EI and EE is provided in appendix F. This includes a full summary of the remodelled EE & EI for the baseline, SHQS period, 2020 and 2050. A summary of each remodelled retrofit example and the applied measures are provided in Appendix G.

4.4.1 Significant software related variations

The re-modelled version of retrofit example study 21G (1999-2007 mid-floor flat) showed a noticeable variation from the result in the consultation document modelled using RdSAP 2005. The post-improvement results presented in the consultation document are considerably greater than were achieved when remodelled; the reasons for this are not clear. Based on Changeworks' knowledge and use of the previous and new versions of RdSAP 2009 (v9.90 and 9.91 respectively), we would not anticipate that using version 9.91 would throw up such major differences unless user defined values were substituted for defaults.

Some changes in the newest version of RdSAP 2009 (v9.91) are straightforward, e.g. the need to enter the thickness of solid walls, check for the presence of dry-lining, the ability to model draughtproofing and the ability to account for location. However, any significant changes in the modelling are most likely to arise from entering the optional additional data that version 9.91 allows ( e.g. actual U-values, insulation specifications, window orientation, etc.), as in some cases these could vary significantly from the default values. As discussed previously, these additional data are optional, and in many cases it is unlikely that these would be collected by the property assessor.

To aid understanding of these issues, a small sample of the examples was re-modelled using the new version of RdSAP, both with and without the optional additional data entered. The findings are outlined in Section 5.2.

4.5 Other findings

Throughout the re-modelling, percentage improvements in CO ₂ emissions and energy usage are similar to those presented in the consultation document. There are however some major differences in the EE and EI results, but mainly for electrically-heated properties. As previously stated this can only be the result of using an older version of RdSAP 2005.

4.5.1 Electrically-heated retrofit examples

For the electrically-heated retrofit examples, the EE ratings are higher and EI ratings lower than those presented in the consultation document. Changeworks compared an alternate model of these retrofit examples generated using a different software programme ^[17] , which resulted in EE ratings that were very close to those presented in the consultation document. There are likely to be numerous reasons for these variations ( e.g. different CO ₂ factors), however most are likely to be the result of using different versions of RdSAP.

4.5.2 Gas-heated retrofit examples

For the gas-heated retrofit examples, the SAP ratings are in most cases lower and EI ratings similar to those presented in the consultation document. As with the electrically-heated retrofit examples, some of these differences may be accounted for by the different versions of RdSAP being used.

The following should also be noted when comparing the gas-heated retrofit examples with the re-modelled versions:

• The consultation document uses a 60% efficient boiler for the baseline. More recent versions of RdSAP are unable to model this, as the assumed efficiency of a wall-mounted boiler is 66% ^[18] . This 66% figure was therefore used for the re-modelling.
• The consultation document uses a 90.9% efficient boiler for the improved heating system. However, there are no such mains gas combi condensing boilers ^[19] on the SEDBUK database so a 90% efficient boiler ^[20] was used for validation.