Renewable Electrical Energy Strategies for Low and Zero Carbon Homes David Cowan, Issa Chaer* and Graeme Maidment London South Bank University, 103 Borough Road, London SE1 0AA * Corresponding Author –
[email protected] ABSTRACT This paper presents the results of a study carried out at London South Bank University into how current electrical power generation and storage technologies might be integrated with advanced energy management systems to achieve zero carbon dwellings to 2016 (implementation of the Code for Sustainable Homes (CSH) Level-6) and beyond. Whereas thermal energy demands of buildings are tending to decrease as they become more thermally efficient, electrical energy demands for dwellings are unlikely to decrease significantly, because the increasing efficiency of electrical equipment and appliances tends to be offset by higher levels of ownership and utilization. Baseline energy demand profiles were developed for a range of typical dwellings built to Part L 2006 building standards [1] and then used to predict the energy demand for similar dwellings if built in 2016 to CSH Level-6 [2,3]. A renewable electrical power generation and storage model was developed and used with the electrical energy demand profiles to determine the overall hourly energy balance for the dwelling and to establish the feasibility of achieving a ‘zero carbon’ performance. A practical design approach, using wind turbine and solar PV generation, together with a battery power storage system, was shown to give net zero carbon on an annual basis, although not on a daily basis during winter months. The inclusion of a fuel cell in the energy model could make good the electrical energy generation shortfalls in the winter months.
Introduction A key concern with renewable energy generation, particularly for electrical power, is that peak generation rarely matches the demand profile and in the case of wind and solar PV energy in particular, the power generation may also be unpredictable, except over relatively long timescales. A significant challenge for the designers of low and zero carbon homes will therefore be not only to generate enough renewable electrical energy on-site (or off-site using private wire connection), but also to deliver it when it is needed.
The national grid meets the demand of all energy users through a combination of scale, diversity (geographical and mix of generation types), frequency management and the use of standby and backup generation capacity and energy storage (e.g. pumped hydro). However, for a single dwelling or small scale development at CSH Level-6, the variability of local renewable energy generation will present challenges in meeting such demands. Combining different technologies such as solar PV and micro-CHP generation may help to smooth the imbalance. However, commercially available micro-CHP systems are fuelled on natural gas, operate with high heat-to-power ratios and do not qualify as renewable energy sources. Also CHP system cannot track fast changing loads and overshoots, so there would still be significant import and export of energy from and to the grid over the 24 hour period unless a local energy stores were used to balance short term differences. This paper presents the results of an investigation into how current renewable electrical power generation and storage technologies might be integrated to achieve low or zero carbon dwellings to 2016 and beyond.
Dwelling Types and Energy Models A range of typical dwelling types were considered, including terraced, semi-detached and detached houses, bungalows and flats. The parameters used for each dwelling are in line with the BRE parameters for their standard dwelling and are listed in Table1. 2
Dwelling Type Dwelling Floor Area m Flat 60.9 End-terrace 78.8 Det-bungalow 67.3 Semi-house 88.8 Det-house 104 Table 1: Floor areas for the selected dwellings The overall heat transfer coefficient (U) values used in the Level-6 BREDEM-12 models were 2 the RAB [4] figures, 0.8 W/m K for windows 2 and doors and 0.15 W/m K for ground floors, walls and roofs. In order to determine the maximum available area for PV panels on the different dwelling types a simple spreadsheet model was
Figure 1, presents the potentially available areas on different dwellings for mounting PV systems. A pitch angle of 45° was used for the pitched roof dwellings. It is also assumed that 70% of a flat roof area is available for PV, 35% of a pitched roof total area and 50% of a vertical surface (the values can be changed by the user). The figure shows that for all dwelling types except flats it should be possible to 2 design in at least 20m of PV. For a three storey block of 12 flats the available roof area is 2 estimated to be 10-15m . Flat Roof Area Available for PV
Potential PV Panel Area by Building Type Pitched Roof Area Available for PV
Available Area m2
Minimum Vertical Surface Area Available for PV 50 45 40 35 30 25 20 15 10 5 0
Typical hourly and seasonal electrical load profiles were generated using half-hourly data for average dwellings published by PBPower [6] and normalised to the BREDEM-12 data in order to produce typical 24 hour and seasonal electrical load profiles for each dwelling type. An example for the semi-detached house is shown in Figure 2. Jan
Feb (Dec)
Mar (Nov)
May (Sep)
Jun (Aug)
Jul
Apr (Oct)
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00 Period Ending (hrs)
3 Storey, 4 End-terrace DetSemi-house Det-house Flats/ Floor bungalow
Figure 1: Potential PV panel area by building.
Building Energy Demand Modelling Baseline dwelling characteristics were established using the BREDEM-12 model [5] and the standard dwelling types and parameters detailed in the previous section, see Table 2 for the predicted annual energy load for the different dwelling types.
Annual DHW Energy kWh 2,997
Annual Lighting, Pumps, Fans & Appliance Energy kWh 2,122
Annual Cooking Energy kWh (Elec) 668
Total Annual Heat Energy kWh 5,177
Total Annual Electrical Energy kWh 2,790
Total Annual Energy Consumption kWh 7,967
3,298
3,332
2,620
720
6,630
3,340
9,969
3,447
3,119
2,287
686
6,566
2,974
9,540
88.8
4,026
3,368
2,945
747
7,394
3,692
11,087
104
5,127
3,776
3,504
788
8,903
4,292
13,195
Dwelling Floor Area m2 60.9
Annual Space Heating Energy kWh 2,180
78.8 67.3
Figure 2: Seasonal 24-hour electrical power load profile for semi-detached dwelling Appliance Energy Demand Modelling
Building Type
Dwelling Type Flat Endterrace Detbungalow Semihouse Dethouse
approximately double the electrical energy consumption.
Electrical Power Load (kW)
constructed. The model assumes that for flats the available roof area is shared equally between all units. The model also includes a calculation of the maximum vertical surface area that might be available for PV (assuming that only one façade is suitably aligned and the area is calculated using the smaller horizontal dimension - width or depth).
In order to predict the future electrical energy demand for dwellings, it was necessary to determine a method of accounting for the projected improvements in the efficiency of individual appliances and future changes in levels of ownership and utilization. The methodology selected used a combination of data for individual appliances and lighting that is available from the MTPROG website [7], household number trend data from the UK national statistics website [8] and other data determined using BREDEM-12 methods. Three types of data were available; a) Reference data which includes historical data up to 2008 and assumes no change in appliance energy consumption beyond, only changing levels of ownership.
Table 2: Annual energy loads predicted using (BREDEM-12 model - Part L 2006)
b) ‘P1’ projections – the most likely scenario for energy efficiency improvements and levels of ownership.
The above table indicates that the combined annual space heating and hot water heat energy demand for dwellings constructed to Part-L-2006 building standards is
c) Earliest Best Practice scenario (EBP) which assumes that every possible improvement in product efficiency would be implemented at the earliest opportunity
The weather data downloaded included hourly solar azimuth and elevation angles and the three components of solar radiation (direct, diffuse and global horizontal) and temperature. Details and equations for calculating the power output of the PV module can be found in Cowan [9]. Although maximum output will always be achieved for a panel that is pointing due south, azimuth alignment anywhere between east and west can result in an annual energy output of between 80% and 120% of that available from a horizontal (flat) module, provided that the elevation is optimised for the azimuth. Azimuth and elevation can be varied within the spreadsheet model to see these effects.
Figure 3, shows the result of modelling annual electrical energy consumption for the baseline and MTPROG derived appliance profiles using scenario P1. This indicates that for the baseline profile (constant appliance numbers but improving efficiencies) the energy consumption decreases steadily with time, whereas the MTPROG profile (increasing numbers of appliances) indicates a peak around 2008 before falling off at a steady rate.
Modelling wind power generation would normally be performed using actual measured data from the site, to a time resolution of 10 minutes or better and over an extended period. However, such information were not available for small and micro-wind turbine installations, so that predictions of power output were taken as estimates only, with a high level of uncertainty. The spreadsheet model calculates the power and energy output for each one hour period. The wind speed value is compared with the wind turbine parameters to determine the operating region for each period. The cut-in wind speed and rated power output were taken from the manufacturer’s data sheet.
Annual Electricity Consumption (% of Year 2000)
The appliance population data and the household number trend data were used to estimate the average level of ownership for each appliance type within each household. The baseline appliance profiles were kept constant (apart from the introduction of whole house mechanical ventilation between 2008 and 2016), in order to isolate the effects of improving electrical appliance efficiencies. The appliance populations derived from MTPROG data indicate that for some appliance types the quantity per household hardly changes between 2000 and 2016, whereas for others, such as computers, it is expected to be more than double (and for set top boxes to increase by 10 folds with respect to the year 2000.
120%
110% 100% 90%
80% Baseline (P1) 70% MTPROG (P1)
60% 1980
1990
2000
2010
2020
Year
Figure 3: Predicted change in household annual electricity demand for two profiles. Significant factors include the trend towards owning larger TV screens and the increasing utilisation of computers. Also, the introduction of whole house mechanical ventilation between 2008 and 2016 offsets some of the efficiency improvements in other appliances. Renewable Electrical Energy Generation and Dwelling Energy Management (2008 Scenario) A spreadsheet model was generated to compute the electrical power generated from solar PV and wind turbine generators, using historical weather data files from IES software for a calendar year. The power and energy were calculated at hourly intervals.
The wind power generation model permits the user to specify other wind turbine parameters for system designs, although 1.7 to 2m diameter is considered a practical limit for most single household dwellings. The size of turbine is constrained both by practical considerations (mounting arrangements and load bearing characteristics of the building) and planning consent and aesthetic considerations. The assumed maximum blade diameter of 2m, limits the rated power output to around 1 kW. For a multi-occupancy dwelling (flats) turbines of up to 5kW would be feasible, provided that the roof structure was designed to cope with the forces. The energy storage model used was based on a general approach described by Jenkins and Fletcher et al 2008. However, for this purpose the formulae have developed using the concept of energy storage and power flow rather than charge storage and current flow. Domestic Power System Design and Energy Balance Techniques A typical domestic power system that incorporates both renewable energy generation and storage is shown in Figure 4. The renewable energy sources may generate either
AC or DC power, depending on the type of generation. Static sources such as PV generate DC power whereas rotating devices such as wind turbines normally generate AC power (although small wind turbines often deliver their power as a DC voltage and current).
outage. An override during any time period is flagged to the user.
Inverters are used to convert DC power to AC power and are connected directly to the dwelling’s AC mains Bus with bidirectional inverters linking the battery to the AC Bus. The power flow and limiting current are set by the associated charge controller.
Figure 5, presents typical summary result from a modelled wind turbine based on the Heathrow weather data and 1.7m turbine diameter (rated at about 800W), mounted above the ridge of the roof and that there is no loss of power due to turbulence effects. The chart shows a peak output of around 80kWh in February, falling to less than 30kWh in June (the total annual output is 544 kWh). 90
Total Energy kWh
The power system is assumed to be connected to the grid to allow surplus power to be exported, whilst a power shortage or low battery condition can be made good by importing power from the grid. The dwelling load includes all lighting, electrical equipment and appliances and cooking loads, plus the power required for the dwelling’s power system monitoring and control equipment.
2008-Scenario Renewable Electrical Energy Generation and Energy Management
GRID
DC 1 (E.G. PV)
DC
AC
INVERTER A1
DC
INVERTER A2
RE GEN
DC
(E.G. CHP)
AC MAINS BUS
INVERTER/ CHARGE CONTROLLER
AC
DC
B
40 30
0
APPLIANCE & EQUIPMENT LOADS
HEATING & COOKING LOADS POWER SYSTEM CONTROLLER
50
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec LOAD POWER METERING
AC
DC N (E.G. FUEL CELL) INVERTER AN
RE GEN
RE GEN ISOLATION PROTECTION & POWER METERING CIRCUITS
60
10
LIGHTING LOADS
IMPORT/ EXPORT METERING
AC
70
20
(DNO) CONNECTION (G59/1)
SMART LOAD MANAGEMENT INTERFACE
ENERGY STORE (BATTERY)
Figure 4: Typical domestic power system with renewable energy generation and storage The energy model uses nested IF and AND statements to determine which criteria are being met during any period. If the battery charge falls outside limits, or if the charge or discharge rates exceed limits the balance is always made good by the grid If PV is a significant contributor to the overall renewable energy generation it is unlikely that the system will be able to operate independent of the grid, even if the net annual energy balance is zero, due to the highly seasonal variation in PV output. The model therefore includes an option for operation in ‘grid restricted’ mode whereby grid power import is permitted only during overnight off-peak periods, for the purpose of re-charging the batteries. However, since it is possible (depending on battery capacity and daily power demand) that the battery could discharge to the lower limit before the next off-peak period, the model includes an override function that will permit grid import in order to prevent a power
Month
Figure 5: Monthly electrical energy output for 1.7m dia wind turbine (Heathrow 96-97 weather data) The corresponding PV electrical energy generation, using the same weather data, is shown in Figure 6. The data was normalised to 2 a 1m panel using the data for a commercially available polycrystalline PV module and shows the monthly output for three different mounting arrangements; horizontal mounted on flat roof, another mounted on a 45° pitched roof pointing due south and the third using a tracking system to maintain optimum alignment towards the sun (not practical but was included in order to determine the maximum possible output). Tracking PV
Horiz PV
Pitched Roof Mount PV Array
35.0 PV Electrical Energy Generation (1 m2 Panel Heathrow 96-97) data
Electrical Energy kWh/m2
RE GEN
80
30.0
25.0 20.0 15.0 10.0 5.0 0.0
Month
Total PV + Wind kWh
Total Electrical Energy Monthly Demand
Net Monthly Electrical Power (Gen - Demand)
Poly. (Net Monthly Electrical Power (Gen - Demand))
500 400 300
200 100 0 -100
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
75
1,000
60
500
45
0
30 15 24
48 72 96 Time (hours)
120
144
168
Figure 8: Semi-detached dwelling hourly energy profile for 7 day period in January (restricted grid power import) In summer, the energy balance becomes positive and the dwelling exports significant amounts of energy to the grid. Figure 9 shows the dwelling energy profile for a 7 day period in July, which still relies on the battery to meet overnight power demand, but the renewable energy generators are capable of meeting daytime demand and export power (note that the surplus power is exported only after the batteries have been re-charged, which is generally late afternoon/ early evening). Pgen (=Pren*ŋA) Wh Pload Wh Total grid power import/ export Wh Battery charge/ discharge ΔEb = Pbatt[c]*ŋC+Pbatt[d]/ŋD Wh Battery SOC (end of period) % 2,500.0
110
2,000.0
100
1,500.0
90
1,000.0
80
500.0
70
0.0
60
-500.0
50
-1,000.0
40
-1,500.0
30
-2,000.0
20 0
24
48
72
96
120
144
168
Time (hours)
-200 -300
1,500
0
Power Flow/ Energy (Wh)
Electrical Energy kWh
Total Roof Mount PV Array Electrical Energy kWh
90
-500
Typical Semi-Detached Dwelling Net Monthly Power Demand and Generation for 20m2 PV and 1.7m Dia Wind Turbine (Heathrow 96-97) Wind Electrical Energy kWh
2,000
Battery SOC (end of period) %
The energy balance for any dwelling can be determined by pasting the dwelling 24 hour seasonal electrical energy demand profiles into the spreadsheet, together with the characteristics and size for the selected wind turbine and PV module type. Figure 7 shows the monthly power generation, demand and net energy balance, modelled for a semi-detached 2 house with a 1.7m wind turbine and 20m PV array on a 45° roof, pointing due south. This figure shows the relative contributions from the PV and wind turbine generators together with the monthly energy surplus or deficit and indicates that over a 12 month period the dwelling energy generation and demand are roughly in balance, but there is a shortfall in generated power in the winter and a surplus in summer.
Power Flow/ Energy (Wh)
The results indicate that although the tracking system produced the most annual energy (200kWh), the static panels could still produce significant amounts of energy (152kWh for a pitched roof mount panel and 138 kWh for the horizontal panel). They also show that for a pitched roof mount panel, the electrical energy produced during winter months is significantly higher than for a horizontally mounted panel.
Battery SOC (end of period) %
Pgen (=Pren*ŋA) Wh Pload Wh Total grid power import/ export Wh Battery charge/ discharge ΔEb = Pbatt[c]*ŋC+Pbatt[d]/ŋD Wh Battery SOC (end of period) % 2,500 105
Figure 6: PV electrical power generation for 2 different mounting arrangements (1m panel, Heathrow 96-97 weather data)
Month
Figure 7: Monthly energy balance for semidetached dwelling with PV and wind generation Figure 8 shows a 7 day period in January, with the import of grid power restricted to overnight off-peak periods. This indicates that the renewable energy falls well short of meeting the demand so that most of the power is imported from the grid overnight to charge the battery, which discharges during the day to meet the demand profile. It also shows that for the battery capacity and State of Charge limits used in the model, the battery charge could still drop to its lower limit before the start of the next off-peak period, resulting in an ‘override’ of the restriction to prevent a power outage.
Figure 9: Semi-detached dwelling hourly energy profile for 7 day period in July (unrestricted grid power import) The energy modelling results described above can be applied to any dwelling whose 24 hour seasonal energy demand profile is known and the PV and wind renewable energy generation can be modelled for any turbine size and a PV array of any size, efficiency and orientation. 2016 Scenario – CSH Level-6 Dwelling Power Demands and Technologies The scenario for 2016 used building design standards that would comply with the energy requirements of CSH Level-6, together with projections for domestic electrical appliance and equipment efficiency, ownership and use,
to produce a new set of hourly electrical energy demand profiles, see figure 3 for details. Both profiles indicate a similar percentage reduction in household energy consumption levels. The BREDEM-12 spreadsheets were reworked to determine the annual energy demands for reference dwelling types constructed to CSH level-6 standards, using ‘high efficiency’ dwelling U values. It was assumed that the electrical appliance energy demand would drop by 15%, as predicted above. The total heat and electrical energy demands for the reference dwellings built to Part L 2006 and CSH Level-6 standards are compared in Figure 10. This predicts that heat energy consumption could drop by around 40% for all dwellings, but that electrical energy demand will reduce by only 15%. The predictions for solar PV from an emerging technologies review suggest that the efficiency of polycrystalline PV modules can be expected to improved from around 16% to 20% by 2016, an increase of 25%. A further 5% improvement is expected from increased efficiencies in inverter and power management technologies. Taken together with the predicted 15% reduction in dwelling electrical energy consumption, this would permit the size of the PV array on a semi-detached house to be 2 2 reduced from 20m to around 13m and still achieve annual energy generation/ demand balance. However, this would need to be verified with further modelling. Part L Annual Heat Energy Part L Annual Electrical Energy 10,000
CSH L6 Annual Heat Energy CSH L6 Annual Electrical Energy
Annual Energy Consumption kWh
9,000
8,000 7,000 6,000
5,000 4,000 3,000
2,000 1,000 0 Flat
Endterrace
Detbungalow
Semihouse
Det-house
Figure 10: Annual energy consumption for Part L 2006 and CSH Level-6 dwellings The contribution from improvements in other technologies is more difficult to predict. The efficiency of wind turbines is unlikely to increase significantly and factors associated with specific installations are probably more significant. Hydrogen powered PEM fuel cells and gas powered Solid Oxide fuel cells could become viable within the next few years. Advanced electrical energy storage techniques will only offer significant benefits if they are
more compact, cheaper and more reliable than lead-acid batteries. Smart metering technologies will help to balance the energy demand within dwellings but will only influence total energy demand if they trigger a change in household behaviour.
Conclusions Baseline energy demand profiles were developed for a range of typical dwellings built to Part L 2006 building standards and also used to predict the energy demand for the same building types, built to CSH Level-6 in 2016. Energy models were developed to analyse renewable electrical power generation and storage profiles, using typical weather data and to compare them with demand profiles in order to determine the overall energy balance for the dwelling and whether it could achieve ‘zero carbon’ status. The year 2008 results indicated that whilst for many dwellings it might be possible to achieve net zero carbon on an annual basis, this could only be achieved using a grid connection. The practical system design was based on wind and solar PV generation, together with a battery storage system to balance generation and demand. However, PV could only meet a small fraction of the energy demand during winter months and although the power output from wind turbines is less seasonal, there are significant practical constraints on turbine size and suitability of the building for mounting it. The energy model therefore included an option to restrict grid power import to overnight offpeak periods, recharging the battery energy storage system back to a preset level in order to ensure the availability of power for the next day. The advantages of this approach were that any energy imported from the grid would be at lower cost and would also benefit the grid itself, because of reduced peak load, an important factor when considering large scale distributed energy generation. The scenario for 2016 used building design standards that would comply with the energy requirements of CSH Level-6, together with projections for domestic electrical appliance and equipment efficiency, ownership and use, to predict hourly electrical energy demand profiles. Increased efficiency power generation and storage would significantly improve a dwelling’s overall power system performance (by up to 50% overall), but it would still have to rely on a grid connection if only wind turbine and solar PV generation were employed. The inclusion of a fuel cell in the energy model could make good the electrical energy generation shortfalls in winter from using only
PV and wind turbines and the heat generated could be used for domestic hot water and space heating. 70% of existing housing stock will still exist in 2050, so that the focus on achieving zero carbon dwellings by 2016 should not detract from applying energy efficiency improvement measures and low carbon generation technologies to existing buildings. References 1 ODPM (2006). The building regulations 2000 (2006 edition): L1A, conservation of fuel and power in new dwellings. ODPM, NBS. 2 DCLG (2008a). The Code for Sustainable Homes: setting the standard in sustainability for new homes, Communities and Local Government Publications. 3 DCLG (2008b) "Code for Sustainable Homes: technical guide April 2008." Volume, DOI: 4 RAB (2007). The role of onsite energy generation in delivering zero carbon homes, Renewables Advisory Board. 5 Anderson B. R., C. P. F., Cutland N. G., Dickson G. M., Henderson G., Henderson J. H., Iles P. J., Kosmina L., Shorrock L. D. (2002). BREDEM-12 Model description 2001 update. Watford, BRE. 6 PBPower (2003). Micro generation network connection, BERR. Peacock, A. D. and M. Newborough (2007). "Controlling micro-CHP systems to modulate electrical load profiles." Energy 32(7): 10931103. 7 ProAcumen. (2008). "What if? tool." Retrieved 7/10/08, 2008, from http://whatif.mtprog.com/Default.aspx. 8 ONS. (2008). "Household numbers and projections: Regional Trends 38." Retrieved 8/10/08, 2008, from http://www.statistics.gov.uk/STATBASE/ssdata set.asp?vlnk=7678 9 Cowan, D., Renewable Electrical Energy Strategies for Low and Zero Carbon Homes, an MSc dissertation. 2008/2009.
