Hawthorn Road PassivTerrace: Final Monitoring Report for TSB. Ben Croxford. March 2013

Transcription

1 Hawthorn Road PassivTerrace: Final Monitoring Report for TSB by Ben Croxford March 2013 Bartlett School of Graduate Studies University College London

2 CONTENTS FIGURES... III TABLES... IV ABBREVIATIONS... V EXECUTIVE SUMMARY INTRODUCTION REFURBISHMENT BUILDING FABRIC INFILTRATION LOW CARBON SYSTEMS Rotex Gas- Solar Unit Mechanical ventilation with heat recovery (MVHR): Maico Aeronom WS Shower waste water heat exchanger (Recoh- vert) Occupant Controls MONITORING METHODOLOGY RESULTS FROM MONITORING GAS CONSUMPTION ELECTRICITY CONSUMPTION Mechanical Ventilation with heat recovery (MVHR) Lighting Monitoring equipment Electricity consuming appliances WATER CONSUMPTION Showers or baths Water consuming appliances ROTEX GAS- SOLAR UNIT Maintenance of Rotex MVHR MONITORING Maintenance of MVHR THERMAL COMFORT MONITORING CARBON DIOXIDE MONITORING MONITORING OF MOISTURE WITHIN THE WALL PASSIVE HOUSE PLANNING PACKAGE (PHPP) LIMITATIONS MONITORING LIMITATIONS DISCUSSION HOW DID THE HOUSE PERFORM OVERALL? HOW IS THE FABRIC PERFORMING? ARE THE LOW CARBON SYSTEMS PERFORMING AS EXPECTED? Rotex gas- solar system MVHR Shower waste water heat recovery unit HOW WELL DID PHPP PREDICT THE FIRST YEARS RESULTS? HOW ARE THE OCCUPANTS USING THE BUILDING, CONCLUSIONS i

3 8. EPILOGUE REFERENCES APPENDICES APPENDIX I: BEFORE AND AFTER CALCULATIONS APPENDIX II: NOTES ON PHPP APPENDIX III: SENSOR DETAILS APPENDIX IV: APPLIANCE DETAILS... 3 ii

4 FIGURES FIGURE 1: THE REFURBISHED HOUSE... 4 FIGURE 2: ENGINEERS DRAWING OF PLUMBING LAYOUT... 7 FIGURE 3: LAYOUT OF SHOWER WASTE WATER HEAT RECLAIM UNIT, (RECOHVERT 2012)... 8 FIGURE 4: CUMULATIVE GAS CONSUMPTION ( ) FIGURE 5: DAILY GAS CONSUMPTION VS DAILY AVERAGE EXTERNAL TEMPERATURE ( ) FIGURE 6: GAS CONSUMPTION BY DAY OF WEEK (2011 DATA) FIGURE 7: GAS CONSUMPTION, 5 MINUTE DATA (2011) FIGURE 8: CUMULATIVE ELECTRICITY CONSUMPTION FIGURE 9: ELECTRICITY CONSUMPTION BY TIME OF DAY; EACH POINT REPRESENTS 30 MINUTES (FOR WEEK OF 8-14 TH AUGUST 2011) FIGURE 10: WATER CONSUMPTION BY DAY OF WEEK (ALL GATHERED DATA FROM FEB DEC 2011) FIGURE 11: WATER CONSUMPTION BY HOUR OF DAY FOR ALL SUNDAYS AND ALL MONDAYS (ALL GATHERED DATA FROM FEB DEC 2011) FIGURE 12: FREQUENCY DISTRIBUTION OF ALL 5 MINUTE PERIOD WATER CONSUMPTIONS BELOW 70L, (FIGURE EXCLUDES FREQUENCY OF 0L BEING CONSUMED, WHICH OCCURRED 80% OF THE TIME) FIGURE 13: GAS CONSUMPTION VERSUS SOLAR RADIATION (AUG- DEC 2011 WHERE AVERAGE EXTERNAL TEMPERATURE >15.5C) FIGURE 14: ROTEX PERFORMANCE ON A SUNNY DAY (7 TH APRIL 2011) FIGURE 15: ROTEX PERFORMANCE ON A CLOUDY COOL DAY (5 TH APRIL 2011) FIGURE 16: TEMPERATURES FROM THE MVHR SYSTEM ALONG WITH EXTERNAL AND INTERNAL ROOM TEMPERATURES ON 14 TH NOVEMBER FIGURE 17: PHOTOS OF MAICO MVHR UNIT MAIN FILTERS, BOTH USED AND REPLACEMENT ONES, MAY 2011 (LEFT), JAN 2012 (RIGHT) FIGURE 18: TEMPERATURES IN LIVING ROOM AND BEDROOM FROM AUGUST TO NOVEMBER FIGURE 19: RELATIVE HUMIDITIES IN LIVING ROOM AND BEDROOM FROM AUGUST TO NOVEMBER FIGURE 20: CARBON DIOXIDE CONCENTRATION AND TEMPERATURE MEASURED IN THE LIVING ROOM, (HOURLY AVERAGES) FIGURE 21: CARBON DIOXIDE AND TEMPERATURE IN THE LIVING ROOM, 16/10/ FIGURE 22: CARBON DIOXIDE AND TEMPERATURE IN THE LIVING ROOM, 1 WEEK AT THE END OF SEPTEMBER FIGURE 23: CARBON DIOXIDE CONCENTRATION BY TIME OF DAY (NOV- DEC 2011) FIGURE 24: AIR CHANGE RATE CALCULATION USING LOG OF CO 2 CONCENTRATION DECAYING OVER TIME (14 TH NOV 2011) FIGURE 25: TEMPERATURE MEASUREMENTS OVER THE YEAR FIGURE 26: RELATIVE HUMIDITY MEASUREMENTS OVER THE YEAR FIGURE 27: ABSOLUTE HUMIDITY MEASUREMENTS OVER 1 YEAR iii

5 TABLES TABLE 1 : U- VALUES OF CONSTRUCTIONS, (FIGURES FROM PHPP MODEL)... 4 TABLE 2: MONITORED VARIABLES, (TICK MEANS MONITORED, ~ MEANS SOME DATA GATHERED, X MEANS NOT MONITORED) TABLE 3 : OVERALL HOUSEHOLD ENERGY CONSUMPTION (PER M 2 ) FOR 10 HAWTHORN ROAD FOR TABLE 4 : OVERALL HOUSEHOLD ENERGY COST (PER M 2 ) FOR 10 HAWTHORN ROAD FOR TABLE 5 : TOTAL CO2 EMISSIONS (PER M 2 ) FOR 10 HAWTHORN ROAD FOR TABLE 6 : PRIMARY ENERGY CONSUMPTION (PER M 2 ) FOR 10 HAWTHORN ROAD FOR TABLE 7: MEAN VALUES OF TEMPERATURE VARIABLES ON 14 TH NOV TABLE 8: SUMMARY THERMAL COMFORT DATA, (JAN 2011 JUN 2012) TABLE 9: SUMMARY OF CO 2 DATA FOR JANUARY 2011, TO JUNE TABLE 10: SUMMARY DATA FROM THE INTRA- WALL MEASUREMENTS, ALL ARE MEANS FROM 17 TH JAN TH JUNE 2012 (FILE: HOBOS IN WALL (VERSION 2).XLSX) TABLE 11: RESULTS FROM PHPP FILES TABLE 12 : BEFORE AND AFTER FIGURES TAKEN FROM FILES SUBMITTED BY EIGHT ASSOCIATES AND ACTUAL METER READINGS... 1 TABLE 13 : CONVERSION FACTORS TAKEN FROM THE PHPP EXTENSION FILE TABLE 14 : ESTIMATED ELECTRICITY CONSUMPTION DUE TO APPLIANCES iv

6 ABBREVIATIONS ACH ATAP CO2 DECC DHW MHP Air changes per hour Anne Thorne Architects Partnership Carbon dioxide Department for Energy and Climate Change Domestic hot water Metropolitan Housing Partnership MVHR Mechanical Ventilation with Heat Recovery PHPP R4tf TRV TSB Passive house planning package Retrofit for the future Thermostatic radiator valve Technology Strategy Board v

7 EXECUTIVE SUMMARY This report covers the first 21 months of occupation of Haringey Passivterrace, an extensive refurbishment of an existing mid-terrace Edwardian property aiming to save 80% of energy consumption and carbon dioxide emissions. The house has been deconverted from two flats, one of which was previously damaged by fire. The building fabric has been upgraded to Passivhaus levels, with wool internal insulation applied to the front facade, and polystyrene based external insulation to the rear facade. Wood fibre insulation was used for the party walls, with lime plaster and Intello membrane used throughout the building to provide a breathing wall construction. A Rotex combined gas-solar unit, with solar thermal panels provides hot water and back-up space heating via 3 small radiators on the ground floor and two small radiators upstairs in the larger bedrooms. A Maico whole house, mechanical ventilation with heat recovery (MVHR) system provides filtered fresh air to all rooms except the kitchen, bathroom and toilets where air is extracted and used to pre-heat the supply air. High performance triple glazed windows have been installed throughout the house, which combined with draught proofing measures reduced the permeability to 2.2 m3/m2 Pascals. During the first year the house has been drying out and the occupants have been learning how to operate and settle into their home. The house has been warm with no air quality issues and the occupants seem happy. Using the most recent data, energy consumption can be calculated for the year to June This covered a cool summer and a cold winter. Space heating is approximately 8% of a typical unrefurbished terrace house, hot water consumption is 76% higher than forecast, electricity consumption is 80% higher forecast. Overall household energy consumption was reduced by 73%, total household cost was reduced by 63%, CO 2 emissions were reduced by 79% total primary energy consumption was well above Passivhaus requirements (120kWh/m2) at 230kWh/m2/yr, but a 71% reduction from a typical house. The Retrofit for the future competition aims to include the investigation of factors that could affect the rollout of major refurbishment programmes across the UK. During this project some issues have been discovered that affect the overall performance of the house and would be important to address in larger programmes. The systems installed are easy to use to achieve comfort, but beneath this apparent simplicity they are actually complex and need understanding to be able to optimally operate them. The different systems can operate in conflict; for example, heating while the window is open. Careful insulation of pipes and ductwork is needed especially on cold side ducts for the MVHR system and the pipework around the hot water cylinder to reduce heat loss. 1

8 Commissioning of the various systems is important and needs someone experienced in this field to make sure they work optimally and that the occupants are happy with the settings. Maintenance of the systems is different to those in most typical homes, clarity is needed on subjects such as who is responsible for cleaning or replacing filters in the MVHR system, and how to get the more complex systems serviced. The achievements of the project team can be considered a success, but there are significant differences between forecast consumption and actual consumption. These differences are due to errors in the initial forecast model, differences in expected operation of the systems, and also due to differences between expected occupant use of energy and actual use of energy within the home. How to get the occupants to use less electrical appliance energy, less hot water and to turn the thermostat down, in a house where the cost of operating is so much less than a typical house, remains an issue. 2

9 1. INTRODUCTION This report gives an overview of the first 18 months monitoring of the TSB Retrofit for the future project, 10 Hawthorn Road. It can be read in addition to the report from Metropolitan Housing Partnership (MHP) that references the whole design team, [MHP 2011]. This report considers the monitoring results and includes both a description of the house prior to refurbishment and a description of what was changed as part of the refurbishment. The UK has enacted legislation to reduce carbon emissions by 80% from 1990 levels, by 2050 [Climate Change Act 2008]. Carbon emissions from the domestic sector account for around 25% of the total UK carbon emissions and of these, about 50% are due to space heating [DECC 2011]. With new construction making a tiny impact on these figures, improving existing housing becomes more important. The Technology Strategy Board, (TSB) launched a competition to help increase the capacity of UK construction industry to carry out domestic refurbishments aimed at reducing carbon emissions. The team led by MHP and Anne Thorne Architects Partnership (ATAP), won funding to carry out such a refurbishment on 10 Hawthorn Road. The overall principle of the refurbishment and of the TSB Retrofit for the future project was to improve the performance of the entire property with a goal to make deep cuts in carbon emissions [TSB 2010]. The house selected by MHP was unoccupied (void) and arranged as two flats; the top flat had been burnt out, with holes in roof space, first floor flooring and in the ground floor. Several of the windows were also boarded up. There is no energy consumption data from before refurbishment that can be used to compare with the refurbished house and no meaningful air change testing was possible prior to refurbishment. The first project meetings of the team were in 2009, and the refurbishment of the house was completed just before Christmas Finishing touches were made around the New Year period and the previously selected occupants moved in during the last weeks of January The general approach taken by the team was to insulate the external envelope to Passivhaus standard; to reduce the infiltration rate to as near Passivhaus standard as possible, and to install suitable low carbon systems to provide heating and hot water with an aim to achieve an overall energy consumption reduction of 80% compared to the original house. The details of the refurbishment are given in the next section. 3

10 2. REFURBISHMENT The house is a late Victorian/Edwardian terrace house typical of London, it now has 3 bedrooms with a bathroom and WC upstairs; downstairs there is a kitchen, living room and extra WC. The total floor area is 109m2, with a total volume of 250m3. Figure 1: the refurbished house The house was originally split into two flats, as part of the refurbishment these were deconverted into a single terrace house, high levels of insulation were added, and high performance, triple glazed, Passivhaus certified windows were fitted. Mechanical ventilation with heat recovery (MVHR by Maico) was installed to extract warm, moist air from kitchen and bathroom and to supply warmed, fresh, dry air to all other rooms, this unit has a small heater battery to provide. A Rotex gas-solar unit was installed to provide hot water and back up central heating, the unit is a gas boiler, and hot water tank combined with solar thermal panels. To complete the project low flow taps, low energy light bulbs and low energy white goods were installed. This chapter covers the refurbishment in some detail and is split into three sections, building fabric, infiltration, and low carbon systems, and each of these is considered separately in later chapters Building fabric Different strategies were used for different parts of the building with the aim of achieving a U-value of 0.15 W/m2.K for opaque parts (roof, wall and floor), and 0.8 W/m2.K for the windows. Actual U-values and constructions of the as-built house are given in the table below. Table 1 : U-values of constructions, (figures from PHPP model) Element Construction U-Value (W/m2.K) Ceiling - main house 350mm Thermafleece PB20, Ceiling rear part of house plasterboard 300mm Thermafleece PB20, Wall interior insulation plasterboard 2 x 100mm layers of Wall external insulation Thermafleece 250mm Jablite PB20, EPS, render 60mm Wall party wall 100mm Gutex woodfibre, lime Floor main house plaster 250mm Jablite, OSB

11 Floor rear kitchen 150mm Jablite, OSB Windows Drewexim triple glazed 0.77 Triple glazed windows with wood frames, reaching Passivhaus standard (U=0.77W/m2.K) were used throughout, from A triple glazed inner door, also to Passiv specifications was used, the outer door was kept to comply with conservation of the external facade of the house. In the house, the internal insulation reduces the available space but surfaces should never drop below 17C so more of the available space would be comfortable than before Infiltration A key part of the refurbishment was to install an air tight layer completely enclosing the occupied areas, on internal walls this was formed by wet plaster; on external surfaces by a vapour permeable, air tight membrane. The membrane used was Intello from and joints were taped with Proclima tape, and sealing glue Orcon F. The membrane is humidity-variable with low diffusion in winter and high diffusion in summer. The requirements for Passivhaus certification is 0.6 air changes per hour at 50Pa pressure difference (ach@50pa). For the slightly less stringent EnerPhit rating the requirement is 1 ach@50pa. Pre-refurbishment the house had large holes due to fire damage, so the infiltration rate before refurbishment is not relevant, however it was measured by Chiltern Dynamics at 16.47m3/m2.h@50Pa, on 14/8/2009. After the first fix of the refurbishment the infiltration rate was measured at 5.29 ach@50pa; after final completion this was reduced to 2.08 ach@50pa. (Note the permeability was measured as 2.53 m 3 /h.m 2 this is an average of pressurisation and depressurisation, test carried out on 17/1/2012 Chiltern Dynamics). A further reduction in permeability was only possible by sealing off the utility room (1.76 m 3 /h.m 2 Further reductions would have required removing the staircase and sealing the wall where it was fixed. The connection of the MVHR system and associated pipework appeared to be where leaks were occurring but these were physically impossible to reach post installation. Since completion, laminated flooring has been fitted and a satellite TV cable has been installed Low carbon systems This section covers the different systems installed in the house; namely a combined gas-solar unit, for DHW and space heating, and an MVHR system for providing sufficient fresh air. A shower waste water recovery unit was also installed Rotex Gas-Solar Unit Domestic hot water (DHW), and wet central heating is supplied from a combined hot water cylinder, condensing gas boiler and solar thermal panels. The unit installed is the Rotex GSU 320E, (see Rotex 2010 for more details). 5

12 The gas boiler is directly fitted to the pre-insulated, hot water tank thus minimising heat losses; along with the cold water fill to the tank, there are two pairs of pipes, one for the solar, the other for the DHW. The central heating and the shower waste water reclaim are taken off the DHW pipe. The details of the system, (from the manufacturer) are below: Gas boiler (4-20kW) Modulating Solar: 2 * Daikin solar flat plates (V m2 = 4.72m2 (min yield 535kWh/m2) 300 litre tank Electrical power consumption = W Estimated heat loss at 60*C: 1.7kWh per day Five small radiators (1 in the kitchen, 2 in the living room, one in each of the main bedrooms upstairs) are connected to the cylinder, along with a central heating pump, these are driven by the Rotex controller, and have thermostatic valves (TRV s) to balance heat loads if needed. The space heating for the house is controlled by the main Rotex controller, see section for more details on controls Mechanical ventilation with heat recovery (MVHR): Maico Aeronom WS250 A Passivhaus certified MVHR unit, with a stated efficiency of 85% (Maico 2010) was installed to preheat incoming air using the heat from extracted air from kitchen and bathroom. The system was balanced according to manufacturers specifications to supply 30.5l/s and extract 30.3l/s at the normal setting, which corresponds to ~109m3/hr, or approximately 0.4ac/h. The unit has a frost protection circuit which is a small electrical heater, and a post-heat exchanger, heater battery which can add a small amount of heat to the air supply (Maico 2010). The large main unit is installed into the utility cupboard in the downstairs toilet. It has 4 silencers, one on each main duct to the unit, these had to be shortened slightly to enable installation, the cold incoming duct is partially insulated. There is a thermostat for the unit in the hallway which controls the post-heat exchanger, heater battery, and a programmer/control unit in the kitchen to control summer/winter mode and ventilation speed. Summer mode bypasses the heat exchanger. The three fan settings are stated to provide 100/150/250m3/hr. Maximum power consumption is 130W. In normal mode the system is extremely quiet. (Purchase cost was 6800, Sept 2010) There are two sets of filters that need maintenance, one needs washing approximately each 3 months, (coarse filter class G4), the other needs vacuum cleaning approximately each 6 months, (pollen filter class F7). Some information on filters is given in EN779:2002. In addition the heat exchanger itself should also be carefully cleaned once a year in soapy water. Maintenance information suggests regular replacement of both the coarse filters and of pollen filters: Maico Manual (2010). 6

13 Shower waste water heat exchanger (Recoh-vert) This quite simple device from Shower-save (2010), is a long waste water tube with a thin copper supply water tube bonded to the outside of it. Incoming supply water is warmed by outgoing waste water and fed back into the shower water thus reducing hot water demand. Its installation demands that the shower is installed on the first floor, or higher, leaving a clear, long straight drop for the waste water tube. The tube is designed so that warm waste water from the shower sticks to the walls of the tube ensuring maximum heat exchange. Figure 2: Engineers drawing of plumbing layout. This product only works while showering and the effect is that it will reduce the amount of hot water used. At any one time there is only 1 litre of water in heat exchanger. Overall the product is quoted as being 61% efficient at 7.5l/s. The unit is estimated by the manufacturer to save >1000kWhrs/yr. In the UK the cost is significant at ~ 1000 including delivery, and installation. A simple payback of at least 20 years can be estimated, with a saving of around 50 per year (based on displacing gas heating at 0.05 per kwh). (Pricing reduced as of August 2012 to ~ 450 excluding VAT and installation) 7

14 Figure 3: Layout of shower waste water heat reclaim unit, (Recohvert 2012) Occupant Controls The unit is set up to provide a minimum of 55C, allowing solar top up to 70C. Maximum benefit from solar occurs if DHW draw off occurs during times of high insolation. The occupants can choose to have space heat using the MVHR system or the wet central heating or both. The MVHR is electric heating and the Rotex is gas-solar. The householders have been asked to not use the MVHR heating. The Rotex is programmed to keep the house at 21C during daytime and 20C at night when set to winter mode. In summer mode the space heating is off. There is no thermostat for the Rotex, the heat provision is based on return flow temperature and was commissioned at installation, thermostatic radiator valves (TRV s) on the radiators can be adjusted by the occupants as necessary. The programme used is Automatic II - details The MVHR can provide a small amount of heat using an electric heater after the heat exchanger, this is a high CO 2 option as grid electricity has a high carbon content, 0.54kg CO 2 /kwh (Carbon 8

15 Trust 2012) and is estimated to be unable keep the house comfortable if the external temperature is below about 0C. There is a thermostat in the hallway that can be adjusted by the occupants and will call for heat if the MVHR system is in winter mode. In summer mode the heat exchanger is bypassed. Manual use of the boost button can increase ventilation when needed. The shower waste water unit has no controls. Occupants can open all windows as required. 9

16 3. MONITORING METHODOLOGY The aims of the monitoring methodology were to comply with the TSB requirements and to inform architects, client and occupants of the performance of the home with some detailed analysis. The monitoring aims to provide answers to the following questions. How did the house perform overall? How is the fabric performing Are the low carbon systems performing as expected? How are the occupants using the building? How well did the Passive House Planning Package (PHPP) predict performance? see section 5.1 for more details. Table 2 shows the different variables monitored during the process, the individual sensors used are listed in Appendix III. The TSB Retrofit for the future project allowed the opportunity to try a new mesh based wireless monitoring system. The solution chosen was Fourier technologies DataNet system (Fourier 2010) with an ultra low power PC to store the data and act as a datalogger (FitPC 2010). All variables are monitored at 5 minute intervals, but data capture for all variables is not necessarily 100%, see section 5.1 for more details. Table 2: Monitored variables, (tick means monitored, ~ means some data gathered, x means not monitored) Variable State Variable State Gas ü Shower waste water, flow heat ü Water ü MVHR, 4 temperatures ~ Electricity ü DHW Main consumption, flow, heat ~ Temp, RH, CO2 - Living room ü Hot water tank temperature ü Temp, RH% - Main bedroom Temp, RH% - Hallway ü û Solar collector, solar flow and return temperatures Temp, RH% inside the front wall and the Temp, RH%, Solar - External ~ MVHR party wall, preheat wool electricity insulation, /frost gutex protection board. consumption Approximately monthly visits were made to take main meter readings manually, photos were taken to date stamp results accurately and to provide a back up method of validating the digital readings from the DataNet system. The monthly visits were also a useful source of anecdotal data. In addition it helped the researcher form a relationship with the occupants that helps them understand the systems better and also helps the researcher understand how they use the systems on a month to month basis. Thanks to the monitoring system selected, it is possible to gather detailed solar data from the Rotex Gas-Solar unit, including collector temperature, solar flow and return temperatures, solar pump operation and cylinder temperature. ü ü û 10

17 Results from the monitoring up to June 2012 are given in the next section. These results are compared to those predicted by the software, Passive House Planning Package (PHPP) before refurbishment began. An update to the design PHPP file has been made to produce an as-built PHPP energy consumption estimate, this is considered in the final section of the results. 11

18 4. RESULTS FROM MONITORING Data up to 26 th June 2012 was used to complete this report. Overall gas and electricity consumption figures are given in Table 3. This table includes estimated existing consumption, estimated proposed consumption, and actual consumption. Total household energy consumption was reduced by 73%, total household cost was reduced by 63%, CO 2 emissions were reduced by 79% and total primary energy consumption was above Passivhaus requirements at 230 kwh/m2/yr, a 61% reduction. Looking in more detail at energy consumption, that used for space heating was reduced by 96% instead of 92%, hot water consumption increased by 76% instead of a forecast reduction of 49%. Electricity consumption increased by 80% instead of a forecast reduction of 65%. Reasons for the discrepancies are proposed in the discussion section. The figures for the existing case (before refurbishment) are based on estimates by Eight Associates, (see Appendix I: Before and after calculations), the proposed, or forecast energy use was made during the design phase of the project and was also an estimate. The actual consumption is based on June 2011 to June 2012 meter readings. All figures are based on a constant floor area, of 108.9m2. The overall household energy consumption (Table 3) was reduced by 73% from the original, estimated before refurbishment consumption. In all cases the actual consumption was significantly higher than the estimated proposed consumption, particularly surprising was the high electricity consumption, which is also double the proposed existing consumption. Table 3 : Overall household energy consumption (per m 2 ) for 10 Hawthorn Road for Existing (kwh/m 2 /yr) Proposed (kwh/m 2 /yr) Actuals (kwh/m 2 /yr) Gas Heating Gas Hot water Electric Total OVERALL Total 668 (100%) 55 (14%) 142 (27%) The following tables, are derived from the energy consumption meter readings in Table 3 using conversion factors as stated in Appendix I: Before and after calculations. Table 4 : Overall household energy cost (per m 2 ) for 10 Hawthorn Road for Existing ( /yr) Proposed ( /yr) Actuals ( /yr) Gas Heating 1, Gas Hot water Electric Total OVERALL Total 2,283 (100%) 320 (14%) 855 (37%) 12

19 The unit cost for electricity is higher than that of gas, and as a result of the higher than expected electricity consumption, total cost per year is 1300/yr, less than existing but much higher than expected, (Table 4). Table 5 : Total CO2 emissions (per m 2 ) for 10 Hawthorn Road for Existing (kgco 2 /yr) Proposed (kgco 2 /yr) Actuals (kgco 2 /yr) Gas Heating Gas Hot water Electric Total CHECK OVERALL Total 138 (100%) 11 (8%) 29 (21%) Overall CO 2 emissions from the house are almost 80% less than the original this is less than the 92% reduction forecast (Table 5). As mentioned previously, both gas and electricity consumption are much higher than expected. Table 6 : Primary energy consumption (per m 2 ) for 10 Hawthorn Road for Existing (kwh/m 2 /yr) Proposed (kwh/m 2 /yr) Actuals (kwh/m 2 /yr) Gas Heating Gas Hot water Electric Total OVERALL Total 805 (100%) 88 (11%) 230 (29%) Primary energy consumption considers how much energy is required to generate the electricity used and also to deliver the gas to the house, (see conversion factors used in Appendix I. The ambitious forecast reductions in primary energy consumption were not achieved with overall primary energy consumption almost twice that of Passivhaus specifications (120kWh/m2) with the actual totals coming in at 230kWh/m 2 (Table 6). The next sections consider these data in more detail with reasons for the discrepancies seen being proposed in the discussion section Gas consumption Gas is used to heat the hot water and also to provide space heating when necessary, there is no gas cooking in the house. Hot water is required all year, space heating only in the coldest months so as expected gas usage is highest in the cold months. Looking at annual cumulative gas consumption, see Figure 4, shows 4 trend lines; Winter 1: where gas use is increasing fastest, Summer 1: gas consumption slows, Winter 2: increased consumption but not at the same rate as Winter 1. 13

20 Summer 2: similar gas consumption to Summer1 The following average figures are derived from least squares regression from the points on the graph indicated by the 4 lines, reading from the left hand side the first line applies to Winter1, then Summer 1, then Winter 2, and Summer 2, conversion factors (from m3 to kwh) are given in Table 14. Winter 1 consumption rate (heating and hot water, house warming up and drying out) = 48.5 kwh/day Summer1 consumption rate (hot water only) = 15.8 kwh/day Winter 2 consumption rate (heating and hot water) = 40.7 kwh/day Summer 2 consumption rate (hot water only) = 15.7 kwh/day The summer consumption rate is due only to hot water usage as there is no space heating; this rate is reduced on sunny days due to contribution of the solar thermal panels, and will vary due to the quantity of hot water consumed. Figure 4: Cumulative gas consumption (Jan Oct 2012) Daily average gas consumption can be plotted against daily average temperature, see Figure 5. As would be expected, the general trend shows more gas is consumed on cold days than on hot days. Also the use of the solar thermal panels should mean that less gas is consumed on sunnier days. To show this, each point is coloured by one of three daily total solar radiation categories. The threshold for days being Sunny or Dull or inbetween (Mid) was selected to ensure a good spread of points in each category, it is clear that the minimum for Dull days is around 18 kwh whereas for a Sunny day of similar average temperature, the minimum consumption is around 8 kwh. The variation in consumption depends on several variables including hot water consumption and time of hot water consumption. 14

21 Figure 5: Daily gas consumption vs daily average external temperature ( ) The following figures show weekly profiles during winter and summer, Figure 6 shows higher gas consumption on weekends than weekdays in winter, reflecting the longer hours of occupancy and space heating required. In summer the variation of gas consumption by day of the week is not clear. Figure 6: Gas consumption by day of week (2011 data) The underlying data is gathered at 5 minute intervals, see Figure 7, as can be seen gas is consumed irregularly on the two selected days with the winter day having more periods with gas being consumed, consumption depends on heat losses from the following; water drawn from the cylinder, space heating, heat losses from the cylinder itself and associated pipework 15

22 and heat gains from; the solar thermal panels shower waste water heat recovery unit. Because of these factors gas consumption varies 3 fold between two selected days for example, Sunday 12 th December and Sunday 12 th September. In a given 5 minute period the maximum gas burn seen is 0.17m3, which corresponds to an average power of 22kW. The burner is fully modulated and the minimum above zero burn is 0.01m 3, corresponding to an average power of 1.3kW over a 5 minute period. From the Rotex specifications the modulation is from 4-20kW so approximately in line with that seen, close matching is not possible due to the 5 minute monitoring interval. Figure 7: Gas consumption, 5 minute data (2011) Plotting gas consumption versus degree days shows that as expected more gas is used in the coldest months, see xxxxx. 16

23 Figure 8: Gas consumption, 5 minute data (2011) 4.2. Electricity consumption Cumulative electricity consumption appears as a straight line indicating no clear seasonal trend with a steady consumption of around 14.7 kwh per day, equivalent to a constant consumption of ~600W, see Figure 9. Looking in closer detail electricity use varies considerably during a day; with minimum average power of around 250W, and maximum of around 4000W over any single 30 minute period, see Figure 10. Minimum consumption is generally occurring during the very early morning and a maximum consumption generally in the late afternoon or evening. 17

24 Figure 9: Cumulative electricity consumption Figure 10: Electricity consumption by time of day; each point represents 30 minutes (for week of 8-14 th August 2011) Data gathered at six second intervals is available for some of the period, this also indicates a minimum of 250W. Observations around the house suggest there are a significant number of appliances that draw energy when plugged in such as mobile phone chargers and entertainment devices. This high base load could be reduced significantly if those appliances and chargers with high standby use could be switched off for most of the time this could reduce annual electricity consumption considerably. 18

25 Mechanical Ventilation with heat recovery (MVHR) The MVHR unit, MAICO WS250, has a stated energy consumption for each of the three fan levels as follows 1/2/3: 30W/ 50W/ 95W, 100m3/h/ 150m3/h 250m3/h but this depends on pressure drop across the system. At fan level 1, the unit is barely audible, with fan level 2, the system is audible but not loud. The consumption figures increase with pressure drop so could be higher. The system has frost protection circuit and a post-heat exchanger heater battery. Maximum current is 6.9A indicating maximum possible heating to be ~1600W. Not clear from the manual if this is possible or not, (Maico 2010). This will be investigated further during winter 2012 with electricity consumption being monitored in more detail Lighting All lighting throughout the house uses low energy lightbulbs Monitoring equipment Datanet devices from Fourier Ltd are used (Fourier 2010) these communicate wirelessly with their base station which is connected to an ultra-low power PC, FitPC. All Datanet devices require mains power but have battery backup for approximately 3 days. There is a direct output from the Rotex that is monitored by the FitPC. There are two battery powered Hobo H08 loggers used to monitor interstitial condensation Electricity consuming appliances A survey of appliance numbers and average daily consumption still needs to be carried out but is expected to be significantly higher than that predicted by PHPP. See Appendix IV Water consumption Overall water use is 181m3/yr, which averages out at ~500 litres per day. There are 3 permanent residents, so expected water use would be around 450 litres/day, the total measured is about 10% more than expected. This might be explained by occasional extras staying over and that the two children are teenagers! Water use varies by day of week as shown in Figure 11, with Sunday, the day with highest consumption. Looking by hour of day at Sundays and Mondays shows later peaks on Sundays and earlier ones on Mondays. 19

26 Figure 11: Water consumption by day of week (all gathered data from Feb Dec 2011) Figure 12: Water consumption by hour of day for all Sundays and all Mondays (all gathered data from Feb Dec 2011) Figure 13 shows a frequency distribution of 5 minute consumption data, so 1l was recorded cor 2.8% of all 5 minute periods, the notable peaks seen at 5l and 6l are likely to be toilet flushes. The maximum consumption in a single 5 minute period was 132l, but only on 6 occasions in almost a year did the 5 minute consumption exceed 80l. 20

27 3 Histogram Percent Water (l) Figure 13: Frequency distribution of all 5 minute period water consumptions below 70l, (figure excludes frequency of 0l being consumed, which occurred 80% of the time). The water consuming appliances in the house, are considered below, roughly in order of water consumption Showers or baths From the flow meter on the shower waste water reclaim sensor, we can estimate that showers account for about 80m3/yr (42% of total water used in house), which works out at an average daily usage of ~220 litres/day. With a tank temperature at around 55C, and delivery temperature of ~40C, and the Recoh-vert estimated to deliver warmed water at ~30C to the cold side of the mixer, the volume ratio would be ~2:3 between hot supply and cold supply. In terms of quantity of hot water drawn from the tank, I estimate this to be ~90 litres/day of hot water used, with a ratio of 2:3 for hot to cold water. The heat meter on the shower waste water heat recovery unit indicates that the Recoh-vert device saves 586 kwh/year (May11-May12). This equates to ~12% saving on hot water consumption thanks to the device, and an overall saving of 5% of gas consumption. (Flow rate of shower still be measured) Water consuming appliances The following list contains just those appliances using water within the home: Two dual flush toilets in the house, with 6l/3l flushes. One kitchen tap, two sink taps, one bath tap, one shower, all with flow regulators 21

28 Hotpoint ultima wmd962 - washing machine,~69 litres per program. Hotpoint aquarius dwf33 dishwasher, eco-cycle 14l per program Hotpoint ultima tcd980 condensing tumble dryer. (unknown use per program) Rotex gas-solar unit Detailed monitoring data is received from the Rotex at 5 minute intervals, unfortunately this data feed is not very reliable (interruptions are thought to be due to electrical noise), with a total of around 50% percent of data captured. However sufficient data is captured to be useful in understanding the efficiency of the system. An early discovery from monitored data, was that the minimum DHW temperature setting for the cylinder was 70C, this significantly reduces the chance for solar top-up. The setting was altered to a minimum of 55C. As seen in Figure 5, on average the Rotex unit consumes less gas on sunny days reflecting the contribution of the solar thermal panels. Looking in more detail at days with no space heating, in Figure 14 we can see that about half the gas is used on the sunniest days compared to the least sunny days (saving around 10kWh/day). The savings are influenced by the amount of hot water used and the time of day that the hot water is used. (The cut-off of 15.5C is used as this is the figure degree days are calculated to and is generally considered to be the upper threshold for central heating use). Figure 14: Gas consumption versus solar radiation (Aug-Dec 2011 where average external temperature >15.5C) The following figures (Figure 15 and Figure 16) show how the unit is charged on a sunny day compared to that on a cooler, cloudy day. The red and green lines are tank temperature, and they only drop below 50C when water is being drawn off, at this point the gas boiler cuts in and heats the tank back up to 57C. When solar thermal is available (if the collector temperature is high enough) the tank will heat up to 70C. 22

29 Delta Q gains, is an estimate of the heat energy required to provide the observed raise in the tank temperature. The large pulses are from the gas boiler, the smaller ones from the solar thermal collector, the scale is shifted slightly for ease of reading. Note how the gas will operate during the night to always keep the tank at above 50C. Figure 15: Rotex performance on a sunny day (7 th April 2011) On the cooler day the collector is never warm enough to charge the hot water cylinder, more of the larger pulses in Delta Q gains can be seen on the colder day, 23

30 Figure 16: Rotex performance on a cloudy cool day (5 th April 2011) Optimal behaviour would be to only use hot water when the weather was sunny enough to recharge the hot water tank Maintenance of Rotex This system is unusual in the UK and requires specific servicing by trained personnel. Despite frequent reminders to MHP, to my knowledge, the 18 month service had not been carried out by August MVHR Monitoring There are four ducts, paired on either side of the heat exchanger; fresh air in, exhaust air out are on the outside of the heat exchanger, and warm, fresh air into the rooms, and warm, stale air from bathroom and kitchens, are on the inside of the heat exchanger. Four probes have been installed one inside each duct of the system, close to the MVHR unit. The probes are high quality but have been extremely susceptible to electronic noise, so the data capture is disappointing. For example every day in November had at least some MVHR data missing, with most days missing around 25% of the possible data. Due to the number of dropouts in the data, it helps considering the averages for each of the four MVHR probe readings in conjunction with external weather data, living room and bedroom temperatures, see Table 7. The external temperature at the weather station is the coldest, air is drawn into the house and measured in the duct before the heat exchanger as MVHR ext, this air is tempered by the surrounding buildings and also by being brought into the house. The next warmest air is that leaving the heat exchanger (MVHR Out), this is the air from bathroom and kitchen (MVHR3) that has passed through the heat exchanger passing its heat to the incoming air. Next warmest again, is the temperature in the bedroom, which is supplied with warmed air leaving the heat exchanger as MVHR4, note the living room temperature is warmer than the bedroom, the extracted air from kitchen and bathrooms is at the highest temperature, MVHR3. The heat exchanger raises the external incoming air by 10.3C, and lowers the outgoing air by 6C. The outgoing air has much higher humidity (not measured) so can carry more energy. Using the equation from Lowe, Johnston (1997), the temperature efficiency (ET) of an MVHR system is given by: ET = (MVHR4 MVHR Ext)/((Bedroom+Iiving room)/2) MVHR Ext) ET = ( )/( ) = 94% efficiency. On this day the efficiency is higher than the manufacturer specification of 85%. Table 7: Mean values of temperature variables on 14 th Nov Variable External 8.7 MVHR Ext 12.5 Average Temp (*C) 24

31 MVHR Out 19.4 Bedroom 22.3 MVHR Living room 24.3 MVHR Taking the same day in November, the data is shown as time series data in Figure 17, the frequent drop outs can be seen but the underlying temperatures are also discernible. Figure 17: Temperatures from the MVHR system along with external and internal room temperatures on 14 th November Maintenance of MVHR No official service of the Maico unit has been carried out as of August 2012.The washable filters are shown in Figure

32 Figure 18: Photos of Maico MVHR unit main filters, both used and replacement ones, May 2011 (left), Jan 2012 (right). The washable filters are washed occasionally by the householders, the pollen filters have not been changed, and the heat exchanger has not been cleaned (August 2012) Thermal comfort monitoring Temperature and humidity have been monitored in the living room, on the ground floor at the front of the house, and in the main bedroom on the first floor at the back of the house. The house is so well insulated that the heating is designed to keep the whole house at a comfortable temperature all of the time. The ground floor has small radiators in living room and kitchen, the two main bedrooms also have small radiators. Table 8 shows that the living room is in general slightly warmer than the bedroom and is slightly drier also. The standard deviations of both RH% and temperature are small indicating very stable conditions throughout the year. For example, the temperature difference between living room and bedroom was less than 1C for 36% of the time between Jan 2011 and Oct Table 8: Summary thermal comfort data, (Jan 2011 Nov 2013). Element Data points % data capture Mean Temp. *C Std.Dev. Temp. *C Mean RH% Std.Dev. RH% MaxT MinT MaxRH Bedroom data Living room data % % Time % 26

33 In the time series graphs of Figure 19 and Figure 20, the temperature of the bedroom can be seen to follow that of the living room, and the relative humidities appear even more closely correlated. Note the range of humidity is within 40 to 60%RH for the majority of the time. Figure 19: Temperatures in living room and bedroom from August to November Figure 20: Relative humidities in living room and bedroom from August to November

34 The thermal environment is perceived by the occupants as being very comfortable, they were sometimes warm in the summer but were reluctant to open windows as they knew the house was airtight and they didn t want to waste energy. Once told they were able to open windows during the summer there were fewer complaints about overheating. However there are some remaining complaints particularly relating to the room directly above the utility room, this is also the subject of further investigation (Aug 2012), the utility room temperature is hot possibly due to insufficient pipework insulation around the hot water cylinder Carbon dioxide monitoring Carbon dioxide has been monitored at 5 minute intervals in the living room of the home, see Table 9 for summary. There were several points excluded where the sensor data recorded anomalous readings below ambient levels of CO 2, these anomalies were thought to be caused by pressure changes in the house, other datalosses occurred when the sensor was switched off accidentally. Figure 21 shows hourly averages (for clarity) for almost a whole year. The carbon dioxide levels generally stayed between 400 and 1000ppm for the duration with just a few spikes over 1200ppm indicating more people were present in the room. Table 9: Summary of CO 2 data for January 2011, to November 2013 Element Data points % data capture Average SD Max CO % Figure 21: Carbon dioxide concentration and temperature measured in the living room, (hourly averages) Taking one days data for one of the spikes seen above (16 th Oct 2011), Figure 22 shows a rising CO 2 level and a rising temperature level, The rise suggests several people were in the room, the 28

35 sharp fall in CO 2 at the end of the day accompanied by a fall in temperature suggesting the window was opened. Figure 22: Carbon dioxide and temperature in the living room, 16/10/2012. Figure 23 shows the variation of CO 2 over a week, combined with internal living room temperature. The CO 2 remains within the 400 to 1200 ppm range throughout the week. The living room temperature is very stable varying between 25.5 and 23.5 during the week. 29

36 Figure 23: Carbon dioxide and temperature in the living room, 1 week at the end of November Looking in more detail at patterns of CO 2 concentrations over time there are characteristic peaks and troughs seen at different times, during the night the levels drop to a minimum, around 6 or 7am, then a peak around 9am, before dropping again to a low around 3 or 4pm, before rising throughout the evening. Figure 24: Carbon dioxide concentration by time of day (Nov-Dec 2011). Making an assumption that air change rate of the house can be estimated from CO 2 decay rate a log graph can be made, in this case from the evening of the 14 th November 2011, the CO 2 rate decayed exponentially from a peak of around 1200ppm to around 600ppm in about 6 hours. From 30

37 the graph the air change rate is around 0.56 h-1. This corresponds to a fresh air supply rate of 39 l/s, which is reasonably close to the commissioned supply rate for the house of 30.5l/s, see section 2.3.2, the excess can be attributed to infiltration which was measured at 2.31m3/m2 facade, which works out at about 0.1 ach-1. Figure 25: Air change rate calculation using log of CO 2 concentration decaying over time (14 th Nov 2011) Monitoring of moisture within the wall As the internal insulation was being completed in January 2011, there was an opportunity to monitor moisture concentrations within the walls themselves. Two temperature and humidity sensors (Hobo H08 Onset Computers) were plastered into the party wall, and into the front wall of the upstairs front bedroom. The sensors were placed inside of the air tight membrane (Intello) and between two layers of the insulation. In the party wall case, the insulation is Gutex woodfibre board with a lime plaster finish, the front external wall has two layers of sheepswool (Thermafleece) with a layer of intello between them, then Gutex woodfibre board and lime render on top of this. Both wall constructions are breathing walls and should not experience interstitial condensation. The sensors themselves measure relative humidity and temperature so the absolute humidity is calculated from these measurements. Table 10 shows summary data, with the bedroom warmest with a very small standard deviation (SD) in temperature, then the party wall, with the external wall also significantly warmer than external temperatures. Table 10: Summary data from the intra-wall measurements, all are means from 17 th Jan th June 2012 (File: HOBOs in wall (version 2).xlsx) Element Data points % data capture Temp *C SD Temp *C Min Temp *C RH% SD RH% Abs hum (g/kg) SD Abs Hum (g/kg) 31

38 Bedroom data Party wall (Wood fibre) External wall (Sheeps wool) External data % % % % Reviewing the temperatures, see also Figure 26, the bedroom temperature remains very stable, not dropping below 20C all year. The two interstitial temperatures are in general between that of the bedroom temperature and the external temperature for the whole year with the wool insulated external wall always being colder than that of the wood fibre insulated party wall. The party wall is perhaps colder than expected suggesting a cold bridge to the outside rather than a direct connection to the neighbouring property, the sensor is on the edge of the chimney breast so this is possible. The RH% inside the external wall is high and is under further investigation. More detailed analysis is given below. Figure 26: Temperature measurements over the year. 32

39 Figure 27 shows how the party wall dried out over the course of the first few months (light blue), with the sensor drying to the minimum detectable RH% by early May, the external RH% signal (dark blue) is very variable and hides most of the detail of the external wall. However it is possible to see that the RH% of the external wall remains high (red), much higher than that of the bedroom (green) during the whole summer. On investigation there was a problem with the guttering immediately outside the area of external wall where the sensor was placed, this is likely to have been a source of moisture into the wall from outside. The external gutter leaked, allowing a path for water to flow down the downpipe and in along the bracket fixing it to the wall, this leak was mended once, and Sandwood replaced the gutter (October 2012). This corner is colder than ever before due to the internal insulation and it never sees the sun, it has the potential to hold moisture for long periods over the winter. Further investigations with extra sensors placed into the outside brickwork are ongoing. Reviewing the minimum temperatures for the interstitial sensor on the external wall (5.4C), it is clearly possible that given a hard enough winter the 0C isotherm is likely to be well inside of the external surface of the wall, if the wall is wet, frost could damage the brickwork. Figure 27: Relative humidity measurements over the year. Figure 28 shows absolute humidities with strong peaks for the external wall (red), and only dropping by October. In spring the absolute humidity in the wall is seen to rise again. 33

40 Figure 28: Absolute humidity measurements over 1 year Passive House Planning Package (PHPP) The design version of the PHPP file for Hawthorn road has some differences to the as-built file. By going through the design file carefully, some small changes were made, chief among these are the changes in target infiltration from 1.0 ach@50pa, to the measured 2.08 ach@50pa, and also the temperature of the house averages around 23C rather than the expected 21C. Other changes found are given in Appendix II. A detailed as-built file would require more data to complete than is available (August 2012) These significantly affect the predicted space heat demand (+42%), and the total primary energy consumption (+68%) as shown in the table below. Table 11: Results from PHPP files Design version (kwh/m2.yr) Space heat demand Primary energy consumption As-built version (kwh/m2.yr) Some differences have been found in the way that the windows specifications have been input into PHPP, the U-values of the frames are higher than those used in the building. The effect of this change would be minor, about 2kWh/m2.year, and has not been implemented in the table above, but further discussion of window modelling can be found in the discussion. These go some of the way to explaining the discrepancies between designed and actual energy consumptions. Further consideration is given in the discussion chapter. 34