J. Cent. South Univ. (2012) 19: 1629-1638

DOI: 10.1007/s11771-012-1186-z

Space heating and hot water demand analysis of dwellings connected to district heating scheme in UK

R. Burzynski¹, M. Crane², R. Yao³, V. M. Becerra⁴

1. Technologies for Sustainable Built Environments, University of Reading, Reading RG66AF, UK;

2. SSE Utility Solutions, Thatcham RG194AZ, UK;

3. School of Construction Management and Engineering, University of Reading, Reading RG66AF, UK;

4. School of Systems Engineering, University of Reading, Reading RG66AF, UK

? Central South University Press and Springer-Verlag Berlin Heidelberg 2012

Abstract: To achieve CO₂ emissions reductions, the UK Building Regulations require developers of new residential buildings to calculate expected CO₂ emissions arising from their energy consumption using a methodology such as Standard Assessment Procedure (SAP 2005) or, more recently SAP 2009. SAP encompasses all domestic heat consumption and a limited proportion of the electricity consumption. However, these calculations are rarely verified with real energy consumption and related CO₂ emissions. This work presents the results of an analysis based on weekly heat demand data for more than 200 individual flats. The data were collected from a recently built residential development connected to a district heating network. A method for separating out the domestic hot water (DHW) use and space heating (SH) demand has been developed and these values are compared to the demand calculated using SAP 2005 and SAP 2009 methodologies. The analysis also shows the variation in DHW and SH consumption with size of flats and with tenure (privately owned or social housing). Evaluation of the space heating consumption also includes an estimate of the heating degree day (HDD) base temperature for each block of flats and compares this to the average base temperature calculated using the SAP 2005 methodology.

Key words: domestic hot water; space heating; energy consumption; heat demand

1 Introduction

The UK government announced very ambitious targets to reduce greenhouse gas emissions by at least 34% by 2020 and 80% by 2050 against a 1990 baseline [1]. This commitment is spread across all industries including commercial and residential building sector.

According to UK Energy in Brief 2011 [2], the total inland energy consumption (temperature corrected) slightly decreased by 4% from 221.6 mtoe (million ton of oil equivalent) in 1990 to 212.3 mtoe in 2010. The same publication further reveals that in 2010 emissions from the UK domestic sector contributed about 17% of all CO₂ emissions in UK, and although the total CO₂ emissions have fallen by 17% since 1990, the emissions from residential sector have increased by 8% [3]. According to analyses done by P?REZ-LOMBARD et al [4], energy consumption in UK residential buildings accounted for 28% and commercial buildings for 11% of final energy consumption. Furthermore, the UK government published the UK Low Carbon Transition Plan [5], which envisages reducing these emissions to almost zero by 2050. In the light of increasing CO₂ emissions from residential sector, the plan appears to be very challenging for this sector.

To support the policy requirement of further reductions in CO₂ emissions, the UK government raises  requirements in UK Building Regulations Part L [6] to build more energy efficient homes. The developers of new residential buildings are obliged to calculate CO₂ emissions arising from the buildings’ energy consumption, which should not exceed a set maximum emissions threshold. There is a number of methods and software packages available for calculation of energy consumption and related CO₂ emissions. The UK implemented European requirements about the method for calculation of the design heat load in British Standard BS-EN 12831:2003 [7]. Recommendations regarding design of domestic hot water systems have been included in BS 6700:2006 [8]. However, to assess the building compliance with regulations, the developers are obliged to use the Standard Assessment Procedure (SAP) 2005 [9] or more recently SAP 2009 [10] which includes all domestic heat consumption and a limited proportion of the domestic electricity consumption.

In the case of domestic buildings, the above calculations are very rarely verified with real energy consumption data and related CO₂ emissions. This is because the developers are not obliged to do a post occupancy evaluation. According to research performed by GIL et al [11] based on the post occupancy evaluation studies of BORDASS et al [12] and PEGG and KOLOKOTRONI [13], the real-world building energy performance aligns poorly with design estimations. Some nominally low-energy buildings perform no better than similar traditional buildings. Comparison of energy performance simulation results with measured energy consumption also shows significant differences, even when the modelling is done based on as-built construction details and actual building occupancy conditions. The differences between modelled and real-world performance might be caused by many factors. Among the most likely factors, the following can be mentioned: incorrect assumption and too simplistic approach during the energy modelling, faults in construction and insufficient insulation of the building envelope, suboptimal operation of systems, faults and underperformance of installed equipment, different use of the building than envisaged during the design phase. All the above highlights the importance of post occupancy evaluation for further improvement of design and modelling methods and practices if the UK is to achieve the carbon reduction targets set for the building sector.

This situation drove the authors to undertake post occupancy analysis of the heat consumption for a new residential development located in London. The authors also wanted to investigate relationships between type of tenure and the domestic hot water (DHW) and space heating (SH) consumption of a property.

2 Methodology

It is generally acknowledged that consumption of DHW in residential buildings varies widely between individual houses or flats. A recent energy saving trust study [14] confirmed that it is mainly dependent on the number of occupants living in a given property. However, this number is usually unknown during the design phase of the development. Therefore, there is a need to use factors known by designers that could help to normalise hot water consumption used by different properties. These factors are the number of bedrooms and the floor area of a property.

The challenge of data analysis is that the space heating and domestic hot water consumption data are jointly measured by one heat meter. Therefore, there is a need to develop a method to separate these two components.

This method is based on the assumption that the heat consumption during the summer time does not include a space heating component, therefore is solely DHW which can be extrapolated to give an estimate of annual DHW consumption. By deducting this consumption from the total heat consumption, it is possible to calculate heat consumption used for space heating. Further parts of this section describe the details of the method used to analyse the available data set.

2.1 Data set

Scottish and Southern Energy (SSE), a UK energy  supply company, provided weekly readings from heat meters installed in more than 150 social housing flats and 300 privately owned flats located in southeast London. The automatic heat and electricity metering system reads data from the meters on a daily basis and logs the information into a central database. The metering system is not based on counting kW·h pulses but rather logs the actual heat or electricity meter readings. Therefore, the data are assumed to have a high level of accuracy. The same data are used for billing the residents and there have been few data quality issues.

For each flat the following information is available: flat number, floor level, number of bedrooms, floor area, type of ownership, hand over date and weekly meter reading starting usually from the hand over date. The first residents moved into the flats in summer 2009 and the last flats were not occupied until summer 2010.

2.2 Flats

The flats were built between 2007 and 2010 to the Building Regulations Part L 2006 standard. All properties are arranged in three buildings and connected to a district heating scheme (DH). A plate heat exchanger installed in each flat provides instantaneous DHW. The space heating system is equipped with a programmable thermostat and radiators are fitted with thermostatic radiator valves (TRV). This configuration of DH, the controls, DHW and SH (radiators’) temperatures are very similar to a gas combination boiler system. The only significant difference is that the heat exchanger for domestic hot water is sized to deliver 55 kW of hot water, which is nearly twice the output of a gas combination boiler.

2.3 Selecting suitable data set and period of analysis

In this analysis, the statistics are based on a full year of data obtained from flats occupied between 05/04/2010 and 03/04/2011. Flats with occupancy period of less than one year were excluded from the analysis.

The first four weeks were treated as a settling-in period during which both heat and electricity consumption might vary significantly from normal consumption. However, the start date of occupation was not known, therefore, it had to be estimated by analysing available heat data with the following criteria:

1) Occupation start date had to be at least four weeks later than property hand over date.

2) The initial and minimum heat consumption had to be not less than 10 kW·h per week. The minimum consumption had been determined based on average weekly consumption of hot water for one bedroom flat with one occupant.

There were some issues with reliability of provided data. For example, readings for some weeks were not available and some readings were inconsistent with the previous or following reading. In some periods, the heat demand was zero, suggesting that the flats were vacant or the residents were away for extended period of time. In many instances, despite the fact that the flats had been handed over, the readings showed no heat consumption or small electricity consumption, which might suggest that the flats were not occupied yet. Such data were either corrected using linear interpolation, readings were omitted from the analysis or the flats were completely excluded from further analysis.

2.4 Dividing overall heat consumption into hot water and space heating components

As there was no separate meter for measuring space heating and domestic hot water consumption, it was necessary to split these two parameters using the method described below.

To find the period during which all heat consumption can be treated as hot water only consumption, a heating degree-day (HDD) factor was utilised. Periods with a low number of HDD should indicate times when the building and flats do not require space heating and the remaining heat consumption is domestic hot water demand alone. However, before the HDD can be calculated, a base temperature must be defined. This is the external air temperature above which a given building or flat does not need any heating. As this temperature was in practice unknown, a regression analysis of the buildings’ heat load against HDD for different base temperatures were conducted. For this purpose, weekly HDD for different base temperatures has been obtained from the website DegreeDays.net [15]. This website provides HDD data calculated based on mean, minimum and maximum daily temperatures obtained from Weather Underground [16] website using an enhanced version of a set of equations developed by the UK Meteorological Office. More details about this methodology and equations can be found on the UK Meteorological Office website [17].

SAP 2009 [10] assumes a variation in the daily DHW consumption in different months. This variation of consumption was applied when calculating the DHW demand. Based on the measured results, average DHW consumption during nine selected weeks and the variation in daily consumption are presented in Table 1, and the data have been extrapolated to other months of the year.

Table 1 Monthly variation factor of average daily hot water consumption (volume) [10]

Flats with gaps in heat readings or zero heat consumption for more than two weeks out of nine were excluded from the analysis. These gaps are likely to arise from times when residents are away from their properties and hence are not representative of the rest of the year.

In the final step, for each flat, the estimated annual DHW consumption was deducted from the total heat consumption, providing annual space heating consumption.

2.5 Estimation of HDD base temperature for buildings

For each building, the base temperature has been estimated using regression analysis. For this purpose, the square of the Pearson product moment correlation coefficient (R) through the total weekly building heat consumption and the number of weekly HDD as Y and X data points has been used. The analysis has been performed for base temperatures between 9 ℃ and  24.5 ℃ with 0.5 ℃ step. The base temperature for which R is maximised indicates the outdoor temperature at which the building does not need additional heating.

2.6 Estimation of HDD base temperature for flats

Calculations analogous to those described in Section 2.5 were performed for each flat for which the heating data of complete year were available. The results have been presented in frequency distribution charts.

2.7 Analysis of space heating consumption

The annual SH consumption per square metres has been calculated for each flat. The flats have been divided into two groups: mid floor flats and flats with additional heat loss though roof or floor. For each group, the mean

and standard deviation of heat consumption per square metre of flat floor area have been calculated and are presented. For flats for which SH estimates from SAP 2005 were available (social housing only), an analysis of the differences between SAP 2005 estimates and site measured consumption has been conducted. The results are presented on the frequency distribution charts.

3 Results

3.1 Selected sample of flats for analysis

Out of 450 flats, 222 had sufficient number of reliable records to be included in the analysis of DHW and 208 in SH analysis. Many flats were excluded due to occupancy period shorter than one year. In DHW analysis, 100 flats are owned by private owners and 122 by social housing. In SH analysis, 94 flats were owned by private owners and 114 by housing association. Details of SAP 2005 estimations for SH were available only for social housing flats. More details of the analysed sample are presented in Table 2.

3.2 Hot water demand analysis base period

According to SAP 2005, the HDD base temperatures for individual dwellings varied between  9.4 ℃ for middle floor level flats and 11.3 ℃ for flats over an unheated car park and top level flats. However, regression analysis between the whole building heat consumption and number of HDD shows that the highest level of correlation is for a base temperature of 14.5 ℃. The results of further analysis show that for a base temperature of 14.5 ℃, the number of HDD per week did not exceed 3 for period between 28/06/2010 and 29/08/2010. Therefore, this period was selected as the base period for which heat consumption was treated as being for domestic hot water only. The selected period provided a sample of nine weeks of data.

3.3 Domestic hot water consumption

The results of the analysis show that, in general, the DHW consumption per square metre of flat floor area decreases with increasing flat size. However, this decrease is not very consistent. A comparison of the results of the analysis of measured DHW consumption with results of SAP 2005 modelling is presented in Table 3. The DHW data represent heat measured at the input to a DHW plate heat exchanger and therefore include heat losses from the heat exchanger and DHW pipework to taps. However, it does not include losses related to district heating distribution pipework up to the DHW heat exchanger which SAP makes an allowance for. Therefore, the measured values have been compared with the corresponding values from cell 39 and 40 of SAP 2005 and cell 45 and 46 of SAP 2009 worksheets.

For privately owned flats, the average annual DHW consumption varies from 25 kW·h/(m²·a) for a 1-bedroom flat to 18 kW·h/(m²·a) and 17 kW·h/(m²·a) for 2- and 3-bedroom flats, respectively. For social housing owned flats, the average annual DHW consumptions are 16 kW·h/(m²·a) for a 1-bedroom flat, 13 kW·h/(m²·a) for a 2-bedroom flat and 19 kW·h/(m²·a) for 3-bedroom flats.

Comparison of SAP 2005 estimation of DHW consumption with the site measured values shows significant overestimation of the demand by SAP 2005. For example, SAP 2005 estimates the DHW consumption for a 2-bedroom flat of 66.22 m² to be 30 kW·h/(m²·a), whereas the average measured consumption for such a flat was only 15 kW·h/(m²·a) (average for all 2-bedroom flats). This means that SAP 2005 overestimated the consumption by 100%. For a 3-bedroom flat, this overestimation factor is lower (50%), which is still substantial. The relative positions of the average DHW consumption measured on site, estimated by SAP 2005 and SAP 2009 for an average size 1-bedroom flat can also be observed in Fig. 1 and Fig. 2. In these figures, the value of the average consumption should be read from the horizontal axis rather than the vertical one.

Table 2 Number and floor area of flats selected for analysis of hot water demand

When SAP 2009 is used to estimate DHW demand, the results are closer to the on site measured consumption. In this case, the difference between measured DHW consumption and calculated using  SAP 2009 for 1- and 2-bedroom flats would be reduced to 68% and for 3-bedroom flats to 28%. This analysis indicates that the SAP 2009 methodology estimates DHW demand more accurately than SAP 2005.

Table 3 Average annual DHW consumption categorised according to type of ownership and flat size

Fig. 1 Distribution of occurrences for DHW consumption for 1-bedroom flats

Fig. 2 Distribution of occurrences for DHW consumption for 3-bedroom flats

The analysis of the DHW data also reveals a very wide range of DHW consumption by individual flats. The consumption can be as low as 2-3 kW·h/(m²·a) and as high as 61-71 kW·h/(m²·a).

More detailed analysis of data from flats with low DHW consumption shows that these flats have frequent and often regular periods of residents’ absence revealed by weeks of significantly lower heat consumption. Low average electricity consumption also suggests that these flats may have low occupancy. On the other hand, a closer look at the flat with the highest average heat consumption per square metre shows both high heat and electricity consumption, which suggests that this 1-bedroom flat might be more densely occupied than other flats. Comparison of the minimum, average and maximum DHW consumption categorised by property ownership and number of bedrooms has been presented in Fig. 3 and Table 4.

Analysis of the distribution of the average DHW consumption shows that for all three types of flats, the most frequent value is shifted towards lower consumption levels. This can be clearly seen in Fig. 1 and Fig. 2, presenting DHW consumption distribution for 1-bedroom and 3-bedroom flats.

Fig. 3 Comparison of minimum, maximum and average DHW consumption for different sizes of flats

Moreover, Fig. 1 also reveals that the DHW consumption of social housing owned flats is typically toward the lower end of the consumption range, while private flats are more prevalent at the higher end. This is also true for 2-bedroom flats. However, 3-bedroom flats have a slightly different distribution of frequency of occurrence. Three-bedroom social housing owned flats have higher consumption than privately owned flats.

The above observations are quantified in Table 5, which reveals that the average consumption of DHW in privately owned flats is higher than that in social housing flats by 56% and by 38% in the case of 1-bedroom and 2-bedroom flats, respectively. However, in the case of 3-bedroom flats, the consumption in social housing flats is higher than that in private flats by 11%.

3.4 HDD base temperatures of buildings

The average base temperature calculated using data from SAP 2005 (cell 79) for the first building with social housing owned flats is 9.92 ℃ and for the second is  9.80 ℃. Unfortunately, similar data are not available for the third building with privately owned flats. Nevertheless, regression analysis of the heat consumption reveals that the real HDD base temperatures are much higher and reach 14.5 ℃ for all three buildings. Figure 4 presents correlation factors R² for different base temperatures. It is also worth noticing that the correlation factors for presented buildings are high, indicating very high correlation between building heat consumption and heating degree days for base temperature of 14.5 ℃.

Table 4 Comparison of minimum, maximum and average of DHW consumption for different sizes of flats and types of ownership

Table 5 Domestic hot water consumption comparison

Fig. 4 Correlation factor R² for different HDD base temperatures

3.5 HDD base temperatures of flats

The base temperatures for individual flats vary significantly. The average HDD base temperature calculated based on the HDD base temperatures for individual flats from all buildings is 15.1 ℃ with the median at 14.5 ℃ and a standard deviation of 3.8 ℃. The frequency distribution of HDD base temperatures reveals wide variation between individual flats.

The correlation factor R for more than 75% of flats indicates strong or very strong correlation, and a further 19.2% moderate or higher moderate correlations is presented in Fig. 5. This shows that calculated HDD base temperatures are highly accurate.

As some flats have more external surfaces than others, for more detailed study, all flats were divided into two groups. The Group 1 comprises flats on the first (over unheated space) and top floor and the Group 2 comprises mid floor flats from the first floor (if over a heated space) to the fifth. The average HDD base temperature for first group is 15.1 ℃ with standard deviation of 3.1 ℃ and for the second group 15.0 ℃ with standard deviation of 4.0 ℃.

Fig. 5 Frequency distribution of correlation factor R for individual flats: (a) Very weak correlation; (b) Weak correlation; (c) Moderate correlation; (d) Higher moderate correlation;   (e) Strong correlation; (f) Very strong correlation

The frequency distribution of HDD base temperatures for flats for Group 1 and 2 is presented in Fig. 6 and Fig. 7.

There is very small difference between the average HDD base temperature of the both groups of flats. This probably would have been slightly larger (lower average building’s mid floors (Group 2)). The highest average consumption of 38.6 kW·h/(m²·a) is found for the Group 1 in the building with private flats. Much lower consumption of 26.9  kW·h/(m²·a) is found for flats in the same building but belonging to the Group 2. Social housing flats in the Group 1 consume less heat than corresponding privately owned flats but social housing flats in the Group 2 consume average 31.1 kW·h/(m²·a), noticeably higher than privately owned flats in corresponding group. There is also notable variation of SH consumption between individual flats, indicated by the high standard deviation (more than two thirds of the average SH consumption). More details about SH consumption are provided in Table 6 and Fig. 8 with the extremes presented in Fig. 9.

Fig. 6 Frequency distribution of HDD base temperature for top floor flats in all three buildings

Site measured SH consumption has also been compared with SAP 2005 estimated consumption. Similarly to the DHW, this comparison also reveals significant differences between site measured consumption and that estimated using SAP 2005. In contrast to DHW, for SH, SAP 2005 generally underestimates the consumption. The average SH consumption for all flats in social housing is        32.1 kW·h/(m²·a) for which related SAP 2005 consumption is only 22.2 kW·h/(m²·a).

Fig. 7 Frequency distribution of HDD base temperature for mid floor flats in all three buildings

Table 6 Annual space heating consumption for flats of floor area

Fig. 8 Frequency distribution of SH for individual, privately owned and social housing flats

Fig. 9 Extremes of SH consumption for individual, privately owned and social housing flats

The difference between the averages shows 31% of underestimation of SH consumption. However, as it can be seen from Fig. 10, SH for 37.5% of flats of the first group (black bars on the left side of the chart) is significantly underestimated by SAP 2005.

3.6 Overall site and SAP 2005 results comparison

For flats for which full annual heat consumption data were available (social housing flats only), overall heat consumption of DHW and SH have been calculated and compared to SAP 2005 estimates. This reveals that although SAP 2005 overestimates the DHW by 90% and SH underestimates by 31%, the overall heat consumption of all flats included in this statistic is overestimated just by 8%. The results are presented in Table 7.

4 Discussion

A method of assessing annual domestic hot water and space heating consumption through separation of the two out of overall heat consumption has been proposed and applied to analyse differences between buildings’ design estimate and measured operational heat consumption. For this new development, the major issue was careful assessment of each flat’s electricity and heat consumption based on which the start of the occupancy period could be recognised and sufficient number of reliable records were ensured from the summer period. In this method, a significant uncertainty is the assumption that for selected flats, the DHW consumption pattern is representative for the whole year and can be used to extrapolate available data to other months of the year. However, not having data from separate DHW meter makes it impossible to verify this assumption. Taking the above into account, it can be concluded that this method is useful for estimation of the split between hot water and space heating consumption for aggregated heat consumption of a building or groups of flats. However, it might be unsuitable for validation or verification of results obtained from modelling with SAP and other models for individual flats.

The results of the analysis suggest that the annual DHW consumption seems to be overestimated by SAP 2005. However, the improved formula for calculation of hot water consumption introduced in SAP 2009 significantly reduces this overestimation.

The results also show some significant differences in hot water consumption between flats occupied by social housing and privately owned flats residents. The DHW consumption is higher in private 1- and 2-bedroom flats by 55% and 38%, respectively. Only in 3-bed social housing flats, the average consumption is higher than that in private flats, only by 11 %.

Heating degree days base temperature for whole buildings (~14.5 ℃) and for individual flats appears to be significantly higher than that estimated by SAP 2005 (~10 ℃). This might have been caused by worse energy performance of building envelope, including poorer airtightness and higher U-values of envelope elements. This would result in higher space heating consumption. The accuracy of the calculated HDD base temperature has been confirmed by high correlation factors between heating degree days and weekly space heating consumption.

Fig. 10 Frequency distribution of differences between SAP 2005 estimated and site measured SH for individual, social housing flats
 

Table 7 Total annual site measured DHW and SH consumption compared to SAP estimates

Analysis of space heating consumption clearly shows that flats located on the top floor and the first floor over the unheated car park have higher average heat consumption (36.9 kW·h/(m²·a)) than those located on middle floors (29.2 kW·h/(m²·a)). However, the spread of average SH consumption of different flats shown by high value of standard deviation factor is wide.

5 Conclusions

1) The analysed buildings underperform in the area of space heating but consume much less energy for supply of domestic hot water. In this way, the results verify the argument that the real-world performance of buildings is often significantly different with what modelling and designs envisaged. Therefore, making post occupancy evaluation mandatory would provide very valuable feedback for designers and building developers, helping them to improve the quality of their work. The introduction of mandatory post occupancy evaluation might be a valuable first step, at least in contracts between developers and housing associations.

2) Although the annual heat consumption for the whole building estimated by SAP 2005 was close to the site measured heat consumption (8% difference), the underestimation of space heating demand (winter consumption) and overestimation of hot water demand might have lead the initial heating supply system (particularly the size of the thermal store and CHP units located in energy centre) to be undersized. This aspect might be the next step in the analysis of heat supply efficiency for this development.

3) The work also attempted to improve assumptions for estimation of domestic hot water consumption of flats using two factors: floor area and number of bedrooms.

References

[1] DECC. Climate Change Act 2008 [R]. London: DECC, 2008.

[2] DECC. UK Energy in Brief 2011 [R]. London: Department of Energy and Climate Change, 2011.

[3] DECC. UK Climate Change Sustainable Development Indicator: 2010 Greenhouse Gas Emissions, Provisional Figures and 2009 Greenhouse Gas Emissions, Final Figures By Fuel Type and End-User [R]. London: National Statistics, 2011.

[4] P?REZ-LOMBARD L, ORTIZ J, POUT C. A review on buildings energy consumption information [J]. Energy and Buildings, 2008, 40: 394-398.

[5] GOVERNMENT H M. The UK low carbon transition plan [M]. London: The Stationery Office, 2009: 80.

[6] Building Regulations 2010. Approved Document L1A: Conservation of fuel and power (New dwellings) [R]. London: Office of the Deputy Prime Minister, 2010.

[7] BSI, “BS EN 12831:2003 Heating systems in buildings — Method for calculation of the design heat load” BSI [S].

[8] BS 6700:2006+A1:2009, BSI, 2009. Design, installation, testing and maintenance of services supplying water for domestic use within buildings and their curtilages Specification [S].

[9] DECC. The Government’s Standard Assessment Procedure for Energy Rating of Dwellings (SAP 2005) [R]. Garston: BRE, Department of Eenergy and Climate Change, Version 9.81/9.83, 2009.

[10] DECC. The Government’s Standard Assessment Procedure for Energy Rating of Dwellings [R]. Garston: BRE, Department of Eenergy and Climate Change, Version 9.90, 2010.

[11] GILL Z M, TIERNEY J M, PEGG I M, ALLAN N. Measured energy and water performance of an aspiring low energy/carbon affordable housing site in the UK [J]. Energy and Buildings, 2011, 43: 117-125.

[12] BORDASS B, COHEN R, STANDEVEN M, LEAMAN A. Assessing building performance in use 2: Technical performance of the Probe buildings [J]. Building Research & Information, 2001, 29: 103-113.

[13] PEGG I M, CRIPPS A, KOLOKOTRONI M. Post-occupancy performance of five low-energy schools in the UK [J]. ASHRAE Transactions, 2007, 113: 1/07/2007.

[14] EST. Measurement of Domestic Hot Water Consumption in Dwellings [R]. DEFRA, 2008.

[15] BizEE Software. Heating & cooling degree days-free worldwide data calculation [EB/OL]. [2011-04-28]. www.degreedays.net.

[16] Weather Underground. Weather forecast and report—Long range and local [EB/OL]. [2011-04-28]. www.wunderground.com.

[17] UK_Met_Office. UKCP09 gridded observation data sets—Frequently asked questions [EB/OL]. [2011-04-28]. www.metoffice.gov.uk/ climatechange/science/monitoring/ukcp09/faq.html#faq.

(Edited by HE Yun-bin)

Received date: 2011-07-26; Accepted date: 2011-11-14

Corresponding author: Robert Burzynski; Tel: +44(0)-118-378-8533; E-mail: rbt.burzynski@gmail.com