Domestic and non-domestic energy performance certificates review: supplementary notes

7.Measuring Windows in RdSAP

For most dwellings, there is no requirement under the RdSAP conventions to measure windows sizes when completing a survey, but should there be?

7.1 Default Algorithms

In RdSAP, "window areas are obtained by application of the appropriate equation from Table S4"^[38] of the SAP methodology )see Figure W1 below). These algorithms use three factors to calculate the respective window areas of a property:

• the dwelling type (that is, a flat / maisonette or a house / bungalow),
• the age band of the dwelling,
• and the total floor area of the dwelling.

Figure W1: RdSAP window algorithms (Table S4 from RdSAP 2012 version 9.93 (19 November 2017))^[39]

Within the RdSAP methodology, window areas are deemed to fall into one of five categories:

• typical
• more than typical
• less than typical
• much more than typical
• much less than typical

While these algorithms calculate the typical window area for the dwelling, Appendix S states:

"The window areas calculated using Table S4 are to be reduced by 25% if it is assessed as being less than typical for the age and type of property, and increased by 25% if assessed as being more than typical for the age and type of property."^[40]

Additional guidance on window area calculations in Appendix S notes:

"Typical applies if the surface area of the glazing in the dwelling is essentially as would be expected of a typical property of that age, type, size and character. Even if there is slightly more or less glazing than would be expected, up to 10% more or less.

More than typical applies if there is significantly more surface area of glazing than would be expected (15%-30% more), perhaps because there is a large sun room or numerous patio doors have been added.

Less than typical applies if there is significantly less glazing than would be expected. This is rare as homeowners tend not to take out windows, but a property may have an unusual design with few windows."^[41]

If the window area in the dwelling is assessed as being 'much more than typical' or as 'much less than typical' (emboldment added), "the total window area should be obtained from measurements of each individual window"^[42] … along with:

• the glazing type (i.e., whether single, unknown age, double glazed pre-2003, double glazed 2003 onwards, triple glazed, or secondary glazed)
• the frame type and glazing gap in pre-2003 or of unknown age uPVC double glazed windows
• the U-value, if known
• a note of whether it is a window or roof light;
• a note of whether it is located in the main dwelling or an extension,
• its orientation

Where window areas are measured and entered into RdSAP, the results from the default algorithms are overwritten by the actual measurements.

So, within RdSAP, window areas in a dwelling do not have to be measured if they are assessed as typical, more than typical or less than typical. Except that nowhere in the RdSAP methodology, does it set out what is 'typical'. Effectively, the only way for an assessor to confirm whether a window area is typical, more than typical or less than typical, or not, is to measure all the window areas, aggregate the results, calculate the typical window area from the appropriate Table S4 algorithm, and compare the two. In other words, the assessor needs to measure the windows to confirm that the assessor does not need to measure them – Catch 22! The assessor is not helped in this process by the software, as the authors of this report are not aware of any currently available approved RdSAP software program that assists the assessor in this process by displaying what the typical window area would be for a given dwelling type, age and total floor area.

The question then is, how appropriate / accurate are the RdSAP window area algorithms Scotland? It would appear, not very accurate.

7.2 Comparative study of RdSAP Default Measured Window Areas

In 2009, Alembic Research was commissioned by the Scottish House Condition Survey team of the Scottish Government to prepare a background paper on the appropriateness of the RdSAP window area algorithms for Scotland^[43]. Alembic Research had previously completed full energy audits of 1398 properties across Scotland as part of the research into assessing the impact of the first three years of the Scottish Government's Central Heating Programme on reducing fuel poverty.

These energy audits included identifying the property type, measuring the total floor area in each property, and identifying the age of each property – that is, collecting the three data items used by the RdSAP window algorithms to calculate the typical window area of a dwelling. These energy audit surveys also measured all of the individual window areas in each property (as that was required by the then-NHER software being used in the assessments), so provided the data to allow the areas calculated by the RdSAP Table S4 algorithms to be compared with the actual measured areas for the dwellings. The overall results from this comparison are set out in Table W1 over the page.

What emerged from the overall assessment, and from breaking the data down by the RdSAP age bands, was the consistent over-estimation of window area across all Scottish age bands by the RdSAP algorithms. Overall, more than 62% of all the dwellings compared, were found to have their window area over-estimated by more than 25%. On average RdSAP was found to over-estimate the window areas by 44.9%.

The greatest variation occurred within the pre-1919 dwelling age band, both in terms of the extreme variation in the accuracy of the predicted area, as well as in the overall over-estimation of the predicted window areas. RdSAP predicted 59% more window area per dwelling than measured on site on average. There was also a variability by a factor of 10 between the two extremes of over and under-estimated window area,

RdSAP Age band	% RdSAP under-estimated by >25% of actual m2	% RdSAP estimated within ±25% of actual m2	% RdSAP over-estimated by >25% of actual m2	% Range of RdSAP predicted m2 vs actual window area	Average % difference per dwelling
Pre-1919 (n=358)	4.5	33.8	61.7	43.8 - 496.2%	59.0
1919-29 (n=104)	1.0	40.4	58.6	56.6 – 286.5%	45.4
1930-49 (n=180)	1.7	48.3	50.0	57.5 – 259.4%	30.3
1950-64 (n=430)	0.9	34.3	64.7	64.7 – 311.2%	40.4
1965-75 (n=242)	1.6	31.0	67.4	61.8 – 382.1%	41.1
1976-83 (n=40)	2.5	27.5	70.0	58.0 – 240.2%	50.5
1984-91 (n=41)	0	29.3	70.7	81.7 – 252.7%	50.2
Overall dataset (n=1398)	2.1	35.6	62.3	43.8-496.2%	44.9

Table W1: Summary of Comparison of RdSAP predicted Window Areas with actual Measured Window Areas in Scotland

The smallest over-estimation of window area occurred amongst the 1930 – 1949 age banding. Even then, the range of variability in the RdSAP predicted window area varied by a factor of 4, and the over-estimation of the window area by RdSAP was 30.3% on average.

When the data was disaggregated between flats and maisonettes, and houses and bungalows, there were differences between the two dwelling categories in the accuracy of the predicted window area. Generally, the RdSAP algorithms predicted considerably more window area amongst houses and bungalows than it did amongst the flats and maisonettes. There was only one age band where this pattern was not repeated, and that was amongst the 1976 – 1983 flats and maisonettes age band. Amongst this age band of flats and maisonettes, 83% had a predicted area that was more than 25% greater than the actual area, compared with only 64.3% of the houses and bungalows (see Table W2).

Age band	Flats and Maisonettes (n=716)	Houses and Bungalows (n=681)
	% >25% less than actual	% within ±25% of actual	% >25% more than actual	Average % difference between predicted and actual m2	% >25% less than actual	% within ±25% of actual	% >25% more than actual	Average % difference between predicted and actual m2
Pre -1919 (F+M n= 239 H+B n=119)	6.3	45.2	48.5	43.3	0.8	5.9	93.3	120.2
1919 – 1929 (F+M n= 63 H+B n=41)	1.6	54.0	44.4	30.3	0	19.5	80.5	68.6
1930 – 1949 (F+M n= 114 H+B n=66)	1.8	64.9	33.3	15.1	1.5	18.2	80.3	62.6
1950 – 1964 (F+M n= 146 H+B n=282)	0.7	55.4	43.9	25.5	1.1	23.4	75.5	48.4
1965 – 1975 (F+M n= 115 H+B n=127)	0	43.5	56.5	31.3	3.1	19.7	77.2	50.6
1976 – 1983 (F+M n= 12 H+B n=28)	0	16.72	83.3	41.7	3.6	32.1	64.3	54.2
1984 - 1991 (F+M n= 23 H+B n=18)	0	43.5	56.5	26.4	1.1	11.1	88.9	80.5
All dwellings (F+M n= 716 H+B n=681)	2.7	50.6	46.8	31.3	1.5	18.8	79.7	65.2

Table W2: Summary of Comparison of RdSAP predicted Window Areas with Actual Measured Window Areas in Scotland – Disaggregated by Flats and Maisonettes, and Houses and Bungalows

Amongst flats and maisonettes, four of the seven age-band groupings had a greater percentage of dwellings amongst those where SAP over-estimated the window area by more than 25% compared against those where the SAP-predicted area was within ±25% of actual window area. Overall, however, the number of dwellings where the SAP-predicted window area was within ±25% of actual window area accounted for a slightly higher percentage of the total sample than did those where SAP over-estimated the window area more than 25%. Overall, RdSAP still over-estimated the window area by 31.3% across all flats and maisonettes.

Amongst the houses and bungalow, the difference between the predicted and the actual window area was much larger. All age bands had more than 60 percent of the dwellings amongst those where the over-estimation of the window area by SAP was by more than 25% compared to those within the ±25% of actual predicted window area. For the pre-1919 houses and bungalows, the over-estimation of the window area by RdSAP was 120.2% on average, that is, RdSAP over-estimated the window area by more than double the actual area. For only one age band of houses and bungalows was the over-estimation by less than 50% (i.e. the 1950-64 age band). Overall, amongst these Scottish houses and bungalows, RdSAP over-estimated the window area by 65.2%.

The sample of dwellings in this comparison was representative of the households that received new heating under the Central Heating Programme in Scotland between 2001 and 2004^[44], both in the private sector as well as the local authority and housing association sectors. It was not representative, nor intended to be representative of the Scottish dwelling stock overall. However, it is a large sample of dwellings. Further, the sample included dwellings from across Scotland, including the three island groups. It included dwellings occupied by owner occupiers, private renters, as well as social tenants. This is a significantly larger sample of dwellings, and a more representative sample across the country than appears to have been used in deriving the RdSAP algorithms in Table S4, which appears to have been derived from surveys of 400 dwellings in Cambridge^[45].

The RdSAP conventions is that assessors should measure individual windows when there is more than 30% more than typical window area, or more than 30% less than typical window area. These results indicate that in Scotland, RdSAP over-estimates the window area across most Scottish dwelling types when broken down by dwelling type and age band by more than 30%. In only three age band by house type groupings (1930-49 flats and maisonettes, 1950-64 flats and maisonettes, and 1984-91 flats and maisonettes) did RdSAP overestimate the window area by less than 30%, and in two of these it was still more than 25% out.

These results indicate that default position for RdSAP assessors in Scotland should not be to assume that the window areas are typical and not measure, but rather, they should assume that window areas are much less than typical compared to what the RdSAP window algorithms would predict, and measure all windows in the dwellings being surveyed, particularly in houses and bungalows.

A very limited assessment of the impact of these differences on the SAP scores for 28 instances^[46], when the SAP assessments were adjusted for actual window areas, indicated a potential impact ranging between -4 SAP points to +3 SAP points, all within the ±4 SAP points error band that SAP uses, with most (i.e. 22 of the 28, 78.6%) falling within ±2 SAP points calculated using the typical window areas.

7.3 Glasgow East End Case Study

During 2013 and 2014, 26 pre-1919 sandstone tenemental flats in the east end of Glasgow were assessed using RdSAP for the purposes of assessing their CO₂ emissions. These flats were going to receive internal wall insulation under a utility funded ECO project, where the contractor received funding on the life-time carbon savings as calculated using the ECO rules published by Ofgem, so maximising the CO₂ savings was deemed crucial to the financial viability of the project. The flats themselves ranged in size from 1-bed bedsits to 2-bedroom flats, and comprised varying internal layouts - some only had one external orientation, while others spanned through the tenement. Initial assessments highlighted that RdSAP over-estimated the window areas for these flats by between 17% and 56% (see Table W3 below). The windows had pre-2003 uPVC double glazing units with 12mm gaps. Overall, RdSAP over-estimated the window area by 33%

Under the RdSAP convention, only 18 of the 26 properties should have had their window areas measured as being more than 30% less than predicted by RdSAP, so classified as 'much less than typical'.

The impact of using the actual window dimensions in RdSAP over the default calculation, whether positive or negative, will in part be determined by the type of glazing and its orientation. The default orientation is East-West in RdSAP for all windows: the greater the window area facing south or in a southerly direction (that is, south east through south west on a compass), the bigger the positive impact. Conversely, if the majority of window area faces north or in a northerly direction (that is, north east through north west on the compass), the impact will be negative.

These tenements were on a south east to north west axis: the windows in the 1-bed bedsits all faced south east; but in the flats running through the tenement, some had more glazing on the south east and some had more on the north west. The differences in the SAP, EI, CO₂ emissions, the Space heating costs and Space heating energy consumption from switching from default window areas to measured window areas is set out in Table W4.

Flat	default window area (m2)	actual window area (m2)	% difference between default and actual m2
F12-18B pre	8.55	4.05	0.526
F13-18B pre	8.42	4.59	0.455
F21-18B pre	9.18	6.21	0.324
F23-18B pre	8.42	4.59	0.455
F31-18B pre	9.18	6.21	0.324
F33-18B pre	8.42	4.59	0.455
FO2-24B pre	8.14	4.59	0.436
F21-24B pre	8.42	4.59	0.455
F22-24B pre	9.67	8.37	0.134
F32-24B pre	9.79	8.37	0.145
F01-32B pre	9.69	7.3	0.247
F02-32B pre	10.5	7.11	0.323
F12-32B pre	8.12	3.56	0.562
F22-32B pre	7.64	3.96	0.482
F32-32B pre	8.12	3.56	0.562
F33-32B pre	8.92	6.21	0.304
F01-34B pre	7.96	4.61	0.421
F21-34B pre	11.61	9.84	0.152
F11-34B pre	11.61	9.84	0.152
F31-34B pre	11.61	9.84	0.152
F12-34B pre	9	6.28	0.302
F22-34B pre	9	6.28	0.302
F32-34B pre	9	6.28	0.302
F11-24B pre	8.41	4.59	0.454
F11-32B pre	8.98	6.48	0.278
F02-18B pre	10.81	9	0.167

Table W3: Comparison of RdSAP default window areas with measured window areas in 26 pre-1919 tenemental properties

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	-1	0	0.1	-2	-51
F13-18B pre	0	0	0	-4	-74
F21-18B pre	-1	-1	0	0	6
F23-18B pre	0	0	0	-4	-74
F31-18B pre	-1	-1	0	1	6
F33-18B pre	-1	-1	0	-3	-54
FO2-24B pre	0	0	0	0	-14
F21-24B pre	0	-1	0	-4	-66
F22-24B pre	0	1	0	-16	-95
F32-24B pre	0	0	0	-3	-76
F01-32B pre	0	0	0	0	-1
F02-32B pre	0	0	0	-5	-95
F12-32B pre	0	0	0	-2	-35
F22-32B pre	0	1	0	-6	-116
F32-32B pre	0	0	0	-5	-100
F33-32B pre	0	0	0	1	30
F01-34B pre	0	1	0	-5	-122
F21-34B pre	0	0	0	-5	-90
F11-34B pre	0	1	-0.1	-4	-83
F31-34B pre	0	0	0	-4	-80
F12-34B pre	0	-1	0	1	7
F22-34B pre	0	0	0.1	1	14
F32-34B pre	0	0	0	1	22
F11-24B pre	0	0	0	-3	-74
F11-32B pre	0	0	0	0	1
F02-18B pre	0	1	0	-4	-93

Table W4: Comparing Energy Performance Indicators from Default Window Areas with those Calculated using Actual Window Area

As can be seen in Table W4, changing from the default window areas to calculated areas has almost no impact on the SAP and EI scores, and very small changes on the CO₂ emissions, space heating costs and space heating consumption.

In Table W5 and Table W6 below, the energy performance indicators are set out showing the impact of installing 100 mm of internal wall insulation in these properties for both the default window areas and the actual window areas respectively. Not all of the external walls in these properties received internal wall insulation because of the additional costs associated with moving kitchen and / or bathroom fittings. Areas excluded were accounted for in these post-internal wall insulation assessments.

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	1	2	-0.1	-19	-424
F13-18B pre	1	2	-0.1	-26	-579
F21-18B pre	2	3	-0.2	-41	-891
F23-18B pre	1	2	-0.1	-26	-579
F31-18B pre	5	5	-0.6	-100	-1138
F33-18B pre	1	2	-0.2	-27	-585
FO2-24B pre	1	2	-0.2	-23	-498
F21-24B pre	1	2	-0.1	-26	-482
F22-24B pre	2	2	-0.2	-40	-1337
F32-24B pre	3	5	-0.4	-68	-1446
F01-32B pre	2	2	-0.2	-33	-735
F02-32B pre	3	4	-0.4	-71	-1337
F12-32B pre	2	2	-0.1	-25	-470
F22-32B pre	1	2	-0.1	-12	-228
F32-32B pre	1	1	0	-10	-225
F33-32B pre	2	3	-0.2	-38	-852
F01-34B pre	1	2	-0.1	-16	-363
F21-34B pre	4	5	-0.5	-85	-1619
F11-34B pre	3	5	-0.5	-80	-1619
F31-34B pre	3	5	-0.4	-78	-1534
F12-34B pre	3	4	-0.3	-49	-1117
F22-34B pre	3	4	-0.2	-53	-1144
F32-34B pre	3	4	-0.3	-52	-1099
F11-24B pre	0	1	-0.1	-12	-278
F11-32B pre	1	3	-0.1	-28	-611
F02-18B pre	3	4	-0.3	-59	-1283

Table W5: Impact of Internal Wall Insulation on Energy Performance Indicators for Default Window Areas

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	1	2	-0.1	-28	-608
F13-18B pre	2	3	-0.2	-41	-910
F21-18B pre	2	3	-0.2	-48	-1048
F23-18B pre	2	3	-0.2	-41	-910
F31-18B pre	5	6	-0.7	-118	-1333
F33-18B pre	2	3	-0.2	-41	-886
FO2-24B pre	2	2	-0.2	-32	-704
F21-24B pre	2	3	-0.2	-43	-798
F22-24B pre	2	3	-0.2	-47	-1552
F32-24B pre	3	5	-0.4	-76	-1607
F01-32B pre	2	2	-0.2	-37	-819
F02-32B pre	3	5	-0.4	-82	-1552
F12-32B pre	2	3	-0.1	-35	-646
F22-32B pre	1	3	-0.1	-26	-482
F32-32B pre	2	2	-0.1	-22	-481
F33-32B pre	2	3	-0.3	-41	-930
F01-34B pre	2	3	-0.2	-29	-658
F21-34B pre	4	5	-0.5	-95	-1809
F11-34B pre	3	5	-0.5	-89	-1802
F31-34B pre	3	5	-0.4	-87	-1705
F12-34B pre	3	4	-0.3	-56	-1255
F22-34B pre	4	4	-0.3	-59	-1273
F32-34B pre	4	5	-0.3	-58	-1208
F11-24B pre	1	2	-0.1	-26	-597
F11-32B pre	1	3	-0.2	-34	-734
F02-18B pre	3	5	-0.3	-67	-1475

Table W6: Impact of Internal Wall Insulation on Energy Performance Indicators for Actual Window Areas

As would be expected, regardless of how the window area is calculated, the impact of fitting internal wall insulation in all of these properties is positive: the SAP scores increase; the EI scores increase; the CO₂ emissions go down, the space heating fuel costs reduce, and the space heating energy consumption drops. However, the impact of these improvements is greater generally where the actual window area is measured (see Table W7).

For 16 of the 26 flats, there is no difference in the calculated SAP score between default window area and actual measured area; there is no difference in the calculated EI score; and there is no difference in the CO₂ emissions (at a scale of resolution of 0.1 of tonne of CO₂ for 17 of the 26 flats. The calculated fuel costs and energy consumption were lower across all 26 flats using actual window area compared to the default window area.

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	0	0	0	-9	-184
F13-18B pre	1	1	-0.1	-15	-331
F21-18B pre	0	0	0	-7	-157
F23-18B pre	1	1	-0.1	-15	-331
F31-18B pre	0	1	-0.1	-18	-195
F33-18B pre	1	1	0	-14	-301
FO2-24B pre	1	0	0	-9	-206
F21-24B pre	1	1	-0.1	-17	-316
F22-24B pre	0	1	0	-7	-215
F32-24B pre	0	0	0	-8	-161
F01-32B pre	0	0	0	-4	-84
F02-32B pre	0	1	0	-11	-215
F12-32B pre	0	1	0	-10	-176
F22-32B pre	0	1	0	-14	-254
F32-32B pre	1	1	-0.1	-12	-256
F33-32B pre	0	0	-0.1	-3	-78
F01-34B pre	1	1	-0.1	-13	-295
F21-34B pre	0	0	0	-10	-190
F11-34B pre	0	0	0	-9	-183
F31-34B pre	0	0	0	-9	-171
F12-34B pre	0	0	0	-7	-138
F22-34B pre	1	0	-0.1	-6	-129
F32-34B pre	1	1	0	-6	-109
F11-24B pre	1	1	0	-14	-319
F11-32B pre	0	0	-0.1	-6	-123
F02-18B pre	0	1	0	-8	-192

Table W7: Comparing Energy Performance Indicators AFTER Internal Wall Insulation from Default Window Areas with those calculated using Actual Window Area

Where the actual window areas are less than the areas calculated by the RdSAP algorithms, there will be bigger savings to be reaped from using actual areas when assessing the impact of the wall insulation – insulated walls will lose less heat than glazed areas. The converse is also true.

When the purpose of the EPC was to produce an asset rating within the A to G banding, being out by 1 SAP point or 1 point on the EI scale may not be noticed. Only in a small number of instances would this change impact upon the SAP banding. However, if the purposes of the EPC are being used for compliance, or being used as the basis to calculate payments associated with carbon savings, then this allows for the EPC / RdSAP to be 'gamed' by contractors or others. That is, they could choose to measure windows when it is in their interest, and choose to use the defaults when they were in their interest. This is not the basis of a 'consistently applied methodology'.

7.4 Mixed Glazing Types

Within RdSAP, under Convention 3.12b, it states "Where a mixture of glazing gaps is present, all window areas should be measured"^[47]. If we are staying with default window calculations, would there be a significant difference of just selecting the window type that is most common? Not in these 26 properties. The difference in the energy performance indicators for 100% pre-2003 uPVC double glazing with a 12mm gap compared with 100% 2003 onwards double glazing is set out in table W8.

From the results, the maximum impact for a complete replacement of the pre-2003 double glazing with 2003-onwards double glazing would be at best +1 on the SAP scale, +2 on the EI scale, a reduction of -0.2 tonnes of CO₂, an annual saving of £20 on the SAP-calculated fuel bill, and annual saving of 375kWh of energy consumption. In the same cases, there is no change in the SAP, EI and CO₂ indicators. Where there is a mixture of units then the actual differences will be even less as they will be somewhere between these two extremes.

This comparison between the results for 100% pre-2003 double glazing with 100% 2003-onwards double glazing was repeated using the actual window areas (see Table W9), and the differences are even less than with using default data for the basis of the comparison.

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	1	1	0	-14	-311
F13-18B pre	1	1	0	-13	-285
F21-18B pre	0	1	-0.1	-13	-279
F23-18B pre	1	1	0	-13	-285
F31-18B pre	1	1	-0.2	-31	-360
F33-18B pre	0	1	-0.1	-13	-275
FO2-24B pre	0	1	-0.1	-10	-233
F21-24B pre	1	1	0	-15	-282
F22-24B pre	0	0	0	-14	-284
F32-24B pre	0	1	-0.1	-14	-290
F01-32B pre	1	1	-0.1	-12	-263
F02-32B pre	0	1	-0.1	-15	-284
F12-32B pre	1	2	-0.1	-17	-372
F22-32B pre	1	2	-0.1	-15	-270
F32-32B pre	1	1	0	-12	-265
F33-32B pre	1	1	-0.1	-13	-282
F01-34B pre	1	1	-0.1	-11	-241
F21-34B pre	1	1	-0.1	-20	-375
F11-34B pre	1	1	-0.1	-19	-373
F31-34B pre	1	1	-0.1	-18	-351
F12-34B pre	1	1	-0.1	-13	-296
F22-34B pre	1	1	0	-13	-294
F32-34B pre	1	1	-0.1	-14	-281
F11-24B pre	0	1	-0.1	-13	-286
F11-32B pre	1	2	-0.1	-13	-296
F02-18B pre	0	1	0	-18	-292

Table W8: Comparing Energy Performance Indicators for 100% pre2003 uPVC Double Glazing with 12mm gap and 100% 2003-onwards Double Glazing for Default Window Areas

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	1	0	-0.1	-6	-144
F13-18B pre	0	0	0	-7	-157
F21-18B pre	1	1	0	-8	-188
F23-18B pre	0	0	0	-7	-157
F31-18B pre	1	1	-0.1	-18	-204
F33-18B pre	1	1	-0.1	-15	-150
FO2-24B pre	0	0	-0.1	-6	-129
F21-24B pre	0	1	0	-8	-155
F22-24B pre	0	-1	0	-12	-261
F32-24B pre	0	1	-0.1	-12	-248
F01-32B pre	1	0	0	-9	-199
F02-32B pre	1	0	0	-7	-127
F12-32B pre	0	1	0	-11	-208
F22-32B pre	1	1	-0.1	-8	-139
F32-32B pre	1	1	0	-5	-121
F33-32B pre	1	1	-0.1	-8	-197
F01-34B pre	1	0	-0.1	-7	-144
F21-34B pre	1	1	-0.1	-16	-314
F11-34B pre	1	0	0	-16	-312
F31-34B pre	1	1	-0.1	-15	-293
F12-34B pre	0	1	-0.1	-10	-208
F22-34B pre	1	1	-0.1	-10	-206
F32-34B pre	1	1	0	-9	-196
F11-24B pre	0	1	0	-7	-158
F11-32B pre	0	1	0	-10	-214
F02-18B pre	0	0	0	-11	-237

Table W9: Comparing Energy Performance Indicators for 100% pre2003 uPVC Double Glazing with 12mm gap and 100% 2003-onwards Double Glazing for Actual Window Areas

The results in Table W10 for these 26 flats indicate only very small differences in the energy performance indicators whether calculated using the default window areas or measured window areas when assessing the differences in energy performance when assuming either 100% pre2003 uPVC Double Glazing with 12mm gap or 100% 2003-onwards Double Glazing. These differences will be even less as the actual split of glazing will be somewhere between the two extremes used here, depending upon the actual split in the different types of glazing.

Flat	SAP	EI	CO2 (tonnes/year)	Space heating cost (£/year)	Space heating consumption (kWh/year)
F12-18B pre	0	-1	-0.1	8	167
F13-18B pre	-1	-1	0	6	128
F21-18B pre	1	0	0.1	5	91
F23-18B pre	-1	-1	0	6	128
F31-18B pre	0	0	0.1	13	156
F33-18B pre	1	0	0	-2	125
FO2-24B pre	0	-1	0	4	104
F21-24B pre	-1	0	0	7	127
F22-24B pre	0	-1	0	2	44
F32-24B pre	0	0	0	2	42
F01-32B pre	0	-1	0.1	3	64
F02-32B pre	1	-1	0.1	8	157
F12-32B pre	-1	-1	0.1	6	164
F22-32B pre	0	-1	0	7	131
F32-32B pre	0	0	0	7	144
F33-32B pre	0	0	0	5	85
F01-34B pre	0	-1	0	4	97
F21-34B pre	0	0	0	4	61
F11-34B pre	0	-1	0.1	3	61
F31-34B pre	0	0	0	3	58
F12-34B pre	-1	0	0	3	88
F22-34B pre	0	0	-0.1	3	88
F32-34B pre	0	0	0.1	5	85
F11-24B pre	0	0	0.1	6	128
F11-32B pre	-1	-1	0.1	3	82
F02-18B pre	0	-1	0	7	55

Table W10: Comparing Energy Performance Indicators between 100% pre2003 uPVC Double Glazing with 12mm gap and 100% 2003-onwards Double Glazing: Actual Window Areas compared to Default Window Area differences

Overall, the results for mixed glazing suggest as an asset rating it does not matter much if there are mixed types of glazing present. Either measure or use defaults across all surveys. Requiring assessors to measure because of the presence of different types of glazing seems an unnecessary demand, and one that an assessor can easily 'game' quality assurance anyway, by only providing evidence of one type of glazing being present in the dwelling.