This is particularly true with photovoltaics. No-one can deny their high price and laughable payback periods – at least on conventional cost grounds. Even the most competitive photovoltaic installation will struggle to get below 30-40 p/kW. On the plus side, photovoltaics capture the imagination like no other clean source of power.
They are high-tech and therefore perceived as mainstream. They come in several forms and can integrate well with both modern and traditional forms of architecture. They are silent in operation, require little maintenance and generate both heat and power with acceptable efficiencies. Perhaps most important: funding agencies like them.
Indeed, they like them so much that the application of photovoltaics is often based on fairly specious environmental and economic grounds. Clearly, some practical installations are needed for the technology to graduate beyond the grand demonstration project or the token gesture.
At a recent expert workshop held at the Science Museum in London, designers with extensive experience of photovoltaics met to discuss the results of some recently completed ETSU-funded design studies. The outcome of their work is being exclusively reported here so that building physicists, developers and designers might learn from their experience.
While several projects were assessed, this article will examine three in detail: a solar boarding house at Haileybury and Imperial Service College, a science and technology centre at Morn Hill near Winchester, and the Yorkshire ArtSpace in Sheffield.
Yorkshire ArtSpace
The Yorkshire Artspace Society building is scheduled for a brownfield site in the heart of Sheffield, a short walk away from the National Centre for Popular Music. The project is backed by Sheffield City Council and partially funded by the National Lottery and the European Regional Development Fund1.
Feilden Clegg architects and consultant Buro Happold devised a simple building containing studio workspaces and workshops for around 60 artists, sculptors and craftspeople, together with a public art studio and gallery space. The building's massing (devised before photovoltaics were proposed) consists of a two-storey element facing the street and a six-storey element facing south-east. A glazed atrium connects the two, with a glazed stair tower on the south-west elevation.
With artists spread over 4000 m2 the internal gains are likely to occur at odd times and peak in particular work areas. The building's internal loads are still likely to be quite modest, with a maximum of 40 W/m2 occurring only in "heavy workspace" areas.
The solar orientation of the site is ideal for photovoltaics. The site slopes to the south, and future developments are unlikely to cause overshading. The design team carried out a solar radiation study using a 3-D computer model which revealed a total annual solar radiation of around 900 kWh/m2, of which around 660 kWh/m2/y occurs between April and September.
The next issue was where the solar panels could be placed for maximum efficiency.
Option appraisal: stair tower
If the original concrete stair tower could be replaced by a curtain walling system hung from a structural frame, the designers could provide 187 m2 of mono-crystalline pvs which would generate around 12 155 kWh/y. The spacing of the laminated cells would be designed to allow some natural light to penetrate the stairs, while ventilation would be provided at the top and bottom of the stairwell.
As the stair was thought to suffer partial overshadowing, the designers considered moving it. However the additional building costs wiped out the slight benefit achieved in increased generation. The total cost of this option was £89 650. The generation cost was put at 37 p/kWh (based on a 20 y life), at relatively low solar efficiency of 65% due to the vertical mounting of the pvs.
The south-easterly roof
The second proposal involved the replacement of four north-facing rooflights on the six- storey block with a single, continuous standing-seam roof. Photovoltaic modules tilted at an optimum 30% would be attached to the roof rather than integrated with it, this being the simplest and cheapest method.
This option involved 112 m2 of pvs, generating 10 640 kWh/y. The installation costs of £67 970 were comparatively low, although here was minimal opportunity to offset building materials costs. The high solar efficiency of 95% led to a generation cost of 32p/kWh.
The facade installation
The base proposal for the building provided elevations of fairfaced structural concrete. The designers proposed to replace this with opaque glass cladding on the south east elevations to support 391 m2 of photovoltaic modules.
This rainscreen would be constructed in front of externally insulated concrete blockwork walls between a series of structural columns, with an open cavity behind the glass. Openings would be left for balconies and windows for daylighting and summer ventilation, while in winter trickle ventilators could draw warm air from the cavity for winter pre-heating.
Although the total output from the 228 modules was a massive 25 415 kWh/y (at a solar efficiency of 65%), the installation cost reflected the high amount of building works, coming in at £194 840. This gave a generation cost of 38 p/kWh.
The cost/benefit analysis
By comparing the building's energy loads the design team concluded that a combination of all three approaches would inevitably exceed the estimated kW load of the building. This meant that the viability of the installation would be compromised as excess power could only be sold back to the grid at around one third of the normal authority selling price.
While combining two of the three schemes would be more viable, the design team was acutely aware of the large difference in conversion efficiency between vertically located photovoltaics modules and those positioned at a more optimal angle of 30°. Around 50% more electrical energy is generated by solar optimised photovoltaic modules.
The design team postulated that redesigning the entire building to optimise roof area for 30° angled photovoltaic arrays would bump up the output to 109 630 kWh/y at a generation cost of 28 p/kWh, although the capital cost penalty would be £613 185.
ArtSpace design lessons
The key lessons to emerge from this study was the role of the local electricity authority, and the importance of offsetting the costs of inert building materials in favour of active elements such as photovoltaic modules.
Feilden Clegg found that designs which involved an offset in building cost made the photovoltaic installation nominally more cost-effective. The stair tower scheme, for example, offset a large cost of the building envelope, so that the net cost per square metre was significantly cheaper than the roof installation. The roof scheme, while cheaper, offset very little of the cost of the building fabric.
The big lesson the design team learned on this project was the cost penalties of feeding electricity back into the grid when the pvs were producing surplus power. This comes in the form of fixed costs related to safety equipment for grid connection, and the variable costs for the monitoring and sale of electricity.
For the latter, Yorkshire Electricity levied three charges: a one-off £10 000 fee for the installation of metering and monitoring equipment, and a monthly charge of £2000 to cover purchase of electricity based on maximum kVa, and monitoring and billing. Combined with the 1.83 p/kWh purchase price offered by Yorkshire Electricity, this meant that the ArtSpace building would need to generate a couple of thousand kWh/y simply to cover the charges.
The design team opted instead for a strategy involving adding or shedding loads. Load adding could include pumped water features, battery storage or backup devices, while load shedding would involve simply disconnecting strings of pv arrays. The capital cost was estimated at between £5-10 000.
Haileybury College
Haileybury and Imperial Service College is a public school located near Hertford. Like many schools of its type, it has expanded to meet the growing needs of secondary education.
Studio E architects and consulting engineer Max Fordham Associates were appointed along with other members of a project team to design a 2000 m2 boarding house extension with strong emphasis on solar energy2.
The £2.8 million budget covered accommodation for 63 boarding girls, a housemistress and her family, and a house matron. This involved a compartmentalised dormitory, individual bedroom and study rooms, a laundry, housemistresses house and sundry ancillary rooms.
Not surprising given the rural site, planning criteria included limits on building height, tree felling and other changes to the landscape. Funding from the government's Energy Technology Support Unit (ETSU) was secured to investigate the project's potential for using photovoltaics.
Among its many requests, the ETSU brief placed particular emphasis on ways in which pv modules could be cooled. The design team decided to deal with this by considering how waste heat from pv cells might contribute to the building's energy requirements. To work out how to do that, the design team set out to determine the most effective means of relieving the heat from the pvs. They also had to consider how the solution would affect the appearance of the building, the client's operational requirements, the nature of the grid connection, and whether additional funding would be necessary.
The mutual dependence of these issues resulted in what the design team call a building integrated photovoltaic/thermal solution, or bipv/t for short.
Options and selection
The design team examined six options:
All these options were scored on a matrix against evaluation criteria of operational needs, efficiency, buildability, etc. This exercise showed the third of these options to be the most viable on issues of accommodation needs, solar efficiency, and cost.
In working up the scheme, the design team had to overcome the limitations of a line of mature trees and a hedgerow. This not only involved an assessment of overshading and the effect this would have on the electrical and thermal output of the photovoltaics, but also questions of tree removal, height reduction and a long-term landscape management plan.
Being a college boarding house the designers concluded that the building will not be in continuous use, occupation being largely during term times in autumn, spring and summer. This sporadic use pattern – and particularly the absence of use during the pv installation's most productive periods – showed that the greatest energy load on the building would be thermal rather than electrical.
The solution lay in a photovoltaic module that combined electrical and thermal collection. Research unearthed a potential product that had been developed at the Eindhoven University of Technology, Holland and is now in the final stages of development by the University and Shell Solar.
The unit consists of an external sheet of glass, an air space, and a pv module with a conducting shell and pipework bonded to its inner face (figure 6). This is backed by a layer of insulation. The liquid cooling of the pv panel both raises the efficiency of the cells as well as providing waste heat. The design team calculated that this could be used for a low pressure mechanical ventilation system (which will revert to a naturally ventilation system in summer).
The design team has calculated that 500 m2 of bipv/t will generate 50 000 kWh/y of electricity and 80 000 kWh/y of heat.
Thermal storage
The average solar gain on south-facing surfaces has been estimated at 125 W/m2 over a six month heating season. The designers have calculated that a 20° south-facing roof (bearing in mind the diffuse component) will provide 62·5 W/m2. Assuming a capture rate of 50% over the 250 m2 area, the pv modules will collect an average 7·8 kW through the heating season, or 11% of the building's peak load. This is worth about 27 500 kWh of heat.
Over the year, 1000 W/m2 is expected to fall on the pv panels, 45% of which can be captured. The designers calculate that 120 W/m2 can be used directly, leaving 330 W/m2 for storage in some form.
Consultant Max Fordham Associates proposes to take the heat straight off the pvs and run pipes through the clay substrate via the building's concrete piles (figure 5). The store itself must be 15 x 15 x 30 m deep giving a volume of 6750 m3. With a heat capacity of 0·83 kWh/m3, the heat capacity of the store is estimated at 5600 kWh/K. To store 82 500 kWh will therefore need a 14·7K swing.
Fordham's calculated the inner core of the heat store will range in temperature from 28 to 42°C, and the outer layer from 18 to 32°C. Heat loss is expected to be 41 000 kWh/y.
Electricity generation
The 50 000 kWh/y of pv-generated electricity will easily meet the building's electrical load. Unusually, the local power supplier, Eastern Electricity, has agreed to credit pv-generated electricity not used by the boarding house against the electricity used elsewhere on the Haileybury estate through other meters.
The only stipulations are that all meters on the estate must be half-hour recording types, and the solar boarding house must have an import/export meter. All the meters will be read on a half-hourly basis. The total consumed, less the electricity exported from the boarding house, will be billed.
Eastern Electricity was able to make this arrangement on the basis that the amount produced by the pvs was small and would never exceed demand on the estate.
This net metering approach means that the school should get the full financial benefit of the power produced by the pv building – a useful precedent for consumers with embedded generators and a separately metered distribution site.
The costs
The capital cost of the 2117 m2 solar boarding house has been estimated at £3.6 million as against £2.9 million for a conventional scheme, although the solar scheme has provided an extra 185 m2 of study bedrooms.
Around £700 000 is for the bipv/t installation, which is estimated to save £4575/y of gas and electricity. This equates to a payback of 152 y.
The scheme has currently reached RIBA Stage E. With funding assistance from the EU Framework 5 programme, the project is due to start on site in May 2000. For the record, the planners have agreed to some coppicing and felling of diseased and elderly trees in return for local replanting.
Morn Hill, Winchester
Morn Hill comprises a mixed development of B1 workplace units, Intech (a science and technology centre for Hampshire County Council) and a 100-bed hotel. A conventional series of buildings on downland adjacent to the A31 had been designed by architect Broadway Malyan in collaboration with Buro Happold and ECD Energy & Environmental3.
The building was inherently low energy from the outset, but energy grants enabled a redesign of the scheme to incorporate photovoltaics. This involved alternative building forms, and different arrangements of orientation and shading.
For various cost and planning reasons, the design team was committed to three, two-storey commercial buildings and three similar sized accommodation blocks.
As with Haileybury, the design team produced a matrix of options which were scored for effectiveness, cost, client needs and buildability, etc. At one point the team considered using a solar tracker to follow the sun, but discovered that it took all the power from the pv just to move the tracker, so that idea was quickly dropped.
Fourteen schemes were evaluated, the last being the preferred solution. The workplace buildings were reoriented from their northwest south-east axis, splayed like the fingers of one hand, to a staggered line of buildings on an east-west axis. This reduced exposure to the north and south and improved visibility.
The electrical load requirements of a typical workplace building of 3420 m2 were calculated at 18 W/m2, made up of 8 W/m2 for lighting and 10 W/m2 for small power. This gave a total building load of 61·56 kW, equating to 12 466 kWh over a typical month.
Using a sawtooth roof on each of the blocks, the design team created space for 126·6 m2 of pvs at a 50° tilt, providing a total of 10 025 kWh over the eight months of the year when an output of at least 1000 kWh was possible. This provided some 8% of the building's power requirements. For the other four months, the design team considered using combined heat and power.
While the sawtooth roof did the job, the marriage with the buildings was judged architecturally unacceptable. In any case, the designers wanted to devise a more integrated roofscape that was visually appealing and bold, although still in context with the site.
Buro Happold devised a solution involving a weaved shell of laminated softwood, which would span 15 m over both buildings and courtyards. The repetitive curved roof was arranged so that north-facing slopes would be left open between buildings to provide natural cross ventilation. The south-facing slopes would occur over the buildings which would be ideal for supporting pvs. The designers saw that pvs could be densely packed on the south-facing roof elements, but peter out in the middle of the courtyards in favour of clear glass. The dipping characteristic over the courtyards would also be ideal for trapping rainwater.
Studies were carried out to optimise pv area and avoid generation exceeding the electrical load of the hotel and three of the workplace buildings (data on the Intech building was not available).
Photovoltaic technology
Various types of pv array were considered. While technically possible, glass/glass arrays were judged too complex to install and would not provide enough insulation. Attention then shifted to a method pioneered in Germany whereby pv cells are laminated in acrylic. This has the advantages of relatively low cost, ease of forming to curvature, and its high amount of recycled material.
The pv cells are married to an American insulating cladding system, where the pv/acrylic laminate replaces the normal outer skin of the cladding. The resulting laminate is highly insulated (as low as 0·56 W/m2), translucent and uv-resistant.
One concern was the product's performance in fire, but a desktop study concluded that it would meet all statutory requirements. The whole composite structure would, however, need to be tested under fire conditions, as would continued watertightness under thermal expansion and contraction.
The design team considered that the roof's repetitive form would permit electrical cabling to arrays which follow the structural grid, with additional logical routing to the plantroom-located inverters.
The costs
So how much did all this cost? The bipv scheme for Morn Hill exceeded the target cost of £900/m2 by around 12%, the entire canopy costing £4.75 million. The design team defends this extra cost (and lack of material cost offset) for the additional shading to workplace buildings, rainwater capture area, weather protection, and its unified appearance.
In the final analysis, there were 5110 m2 of photovoltaics, generating 513 000 kWh/y. This is with arrays at 40° tilt, and a 90·9% operating efficiency. For a 20 year pv life, and an undiscounted costing, power could be generated at 15·8 p/kWh. At a discount rate of 6%/y, the cost would be 27.5 p/kWh.
The technology – what is bipv?
Building integrated photovoltaics (bipv) is the term coined to describe photovoltaic modules designed as part of the material fabric of a building. Integration also implies that the added capital costs for on-site electricity generation can be partially offset by the omission of conventional fabric elements, such as roof tiles, wall cladding and shading devices. Photovoltaic technology can currently be divided into amorphous cells (otherwise called thin-film technology) and mono or poly-crystalline cells. While thin-film technology has yet to make a meaningful impact, the UK market for crystalline cells is dominated by BP Solar and Siemens. Other manufacturers such a Kyocera and Shell are likely to enter the bipv market shortly. A number of manufacturers produce pv modules, formed by laminating pv cells onto glass or plastics. The market leaders include Schücco and Pilkington Solar International. Costs for pvs tend to vary considerably, but £340/m2 for a laminar unit is often quoted, with glass modules costing in the region of £450/m2. That said, recent trends show a net reduction in capital cost of at least 7% each year. Further reading Photovoltaics in building; a design guide Report ETSU S/P2/00282/REP, 1999.Source
Building Sustainable Design
Reference
1'Design study for bipv installation Yorkshire ArtSpace, Sheffield', ETSU report S/P2/00307/REP, 1999. 2'Building integrated design study: solar boarding house, Haileybury and Imperial Service College', ETSU report S/P2/00312/00/00, 1999. 3'Building integrated photovoltaic design study', Broadway Malyan summary report, 1999. The results of the ETSU-funded investigations into photovoltaics will be published in book form later this year.
