Green Week: Checklist for fitting solar panels in social housing

Mark Seaborn, managing director of housing consultancy Pennington Choices in Runcorn, assesses the feasibility of fitting photovoltaic solar panels to social housing stock.

Most social landlords will by now have come across the concept of installing PV on the roofs of their housing stock, in order to reduce electricity consumption from the grid and to take advantage of the government's Feed in Tariff scheme. Many social landlords are now making moves to undertake feasibilities and to commission installations, some with a wary eye on the Government's review of the FiT tariffs in April 2012. While there is a sense of urgency to this, with many commentators expecting tariffs to go down for new systems installed after April 2012, like any other long term capital works project there is also a need to get it right in the first place.

The starting point for any social landlord thinking of installing PV panels is to establish which properties are suitable, what the expected capital costs are and what revenue might be generated. For all but the smallest of social landlords, the starting point is to undertake a desktop study where each property is assessed, via web based imagery such as Google Earth to determine the suitability of the property for a PV installation. Properties need to be:

• Ideally within 90 degrees of South, from East to West

• Have a flat roof or ideally have a pitch of between 30 and 40 degrees

• Have no obvious shading or shadowing issues such as from nearby trees, other properties, other aspects of the roof or a significant chimney.

Not all properties will be suitable for a PV installation, principally due to inappropriate orientation, with proportions of one third of a social landlords housing stock being fairly typical. As a result there is clear need to identify those properties which are appropriate in a quick and cost effective way. Starting from a total stock list, the buildings that might be suitable for a PV installation can be ascertained by a desktop assessment, which while it has some inaccuracies, can quickly and cheaply produce a 'short list' of housing stock for further investigation.

By accessing imagery of the property, typically via Google Earth, but foresee ably via the social landlords asset management system, where elevation photographs are held alongside stock condition data the size of roof can be estimated, either using specialist software or by a visual assessment. The actual size of installation that can be fitted to an individual roof is constrained by the shape and size of the roof, the constraints imposed by planning regulations and the manufacturers fitting requirements. While at first glance this might appear complex, these requirements can be distilled to some simple 'rules' which can then be applied to each roof. For a recent project for a local authority client we constructed a bespoke database using MS Access, making the task of analysing several thousand roofs a manageable and practical task. The roof dimensions and characteristics were input; taking data from Google Earth, the database then calculated the size of PV installation that might be appropriate for each property that met the basic orientation and shading criteria.

The financial feasibility of any particular installation is a reflection of the electricity that it will generate. This can be calculated for each property, by reference to:

• The size of the installation, as derived as described above

• The efficiency of the panels to be installed – different manufacturers quote different efficiency rates

• The expected amount of daylight hours that a property should be exposed to over the course of a year. This can be determined by reference to sun tables which state the average hours of daylight that a property can expect to be exposed too, based on its location in the country. Our experience of advising clients in this area is that most installers simply use the national average, but that there is a significant difference in daylight and hence electricity produced between southern and northern locations, which of course affects the financial feasibility of installations.

• The orientation of the property, relative to South

A simple algorithm will then calculate the amount of electricity that a particular installation should produce. It's important to remember the obvious – this is a forecast, with the level of daylight being the key variable that can change from one year to another. However our experience is that as long as the forecast is calculated accurately, using the specific variables identified above, the actual out turn will be very close to the forecast, certainly over the long term of a number of years.

The 'income' that is derived from an installation is effectively made up of three things:

• The tariff paid for the generation of the electricity. This is measured via a meter, fitted as part of the installation.

• The tariff paid for selling surplus electricity back to the grid. This is typically determined on a 'deemed' basis i.e. an assumption is made about the level of electricity that is provided back to the grid which is currently 50% for domestic installations.

• The saving that the property occupier makes on using the generated electricity and hence avoids buying electricity from the market.

This later point will not necessarily feature in the calculations produced for social landlords in that for domestic properties they won't be the electricity bill payer. Even where the installation is to a scheme with significant communal electricity needs, such as a sheltered housing scheme most service charge mechanisms will result in a reduced cost to the tenants, rather than a return to the social landlord. In our experience the vast majority of social landlords see the provision of the free electricity to their tenants as part of their wider role, with the reduced carbon emissions being another significant motivating factor. We have however come across a small number of landlords who are seeking to recover the 'free electricity' element by:

• Establishing a separate 'utility' business to effectively charge tenants for the electricity produced.

• Increasing the core rent, taking the view that the target rent increases as a result of an increase in the capital value of the property, following the installation of PV.

While these later points may appear somewhat inappropriate, some social landlords have taken the view that as PV is only suitable for a proportion of properties, its not fair that only those tenants receive benefit from the installation and one of the above approaches, allows them to share the proceeds, across their whole property base, rather than just those where installing PV is feasible.

The tariff's that are presently paid heavily incentivise the production of electricity and its consumption locally, on site. This principle is key to government policy and in part is a driver for the desire of government to review (reduce) the tariffs for large scale solar farms where the intention is large scale energy production, which is then passed on to the grid and not used locally. Our experience has been that building uses such as a care and residential homes, offices and leisure centres, all with a high day time electricity consumption can fair disproportionately well in financial feasibility terms. This is because, while they get FiT for all the electricity that they produce they also substantially decrease the electricity bought from the grid, because they use the electricity being generated (in the day) by the PV installation.

To complete the financial feasibility picture it would be usual to produce a long term cashflow forecast, much like social landlords would do for day to day business planning purposes. This cashflow would include the initial cost of installation and the forecasted 'income' over 30 years.

The costs of installation can easily be calculated by reference to a schedule of rates to include items for the material costs of the panels, scaffolding and other fitting costs and mid term maintenance costs. Probably the most interesting aspect of the long term financial forecasting is the assumption made around inflation. Long term, RPI inflation does impact on social landlords, in that the tariff's all increase each year by that amount i.e. a period of high inflation drives up the tariff's paid, but the costs are fixed, hence an overall better financial return. For clients where they are also paying the electricity bill and part of their return is the saving on not buying electricity from the grid, it would be usual to assume an increase in energy prices significantly in excess of RPI. 10%, as opposed to RPI of 3% is not an uncommon assumption in such financial modelling. For the same local authority client mentioned above, all of this analysis work was undertaken in a bespoke database, allowing analysis of each property, quickly and cost effectively. The use of a database tool allows the social landlord, Director of Finance to now properly assess:

• The initial costs of installation

• The level of return for the properties on an individual, archetype or whole stock basis, expressed as a payback period, NPV over say 30 years or rate of return basis.

• Determine which properties are viable for installation from an economic point of view, e.g. where properties have particularly long payback periods / low levels of internal return, perhaps due to orientation, or shading issues or initial cost and hence are not suitable.

Limitations of a Desktop Approach

While a desktop approach as described above, allows large numbers of properties to be assessed, quickly and cheaply it does have limitations. Typically most feasibility studies are conducted using Google Earth imagery and as such the accuracy of the feasibility work will always be limited by this. Practical restrictions include:

• Address labelling is sometimes wrong

• House numbering is sometimes obscured by Google

• Some cul de sacs and estate based enclosed areas have not been captured on the imagery

• The pitch of the roof can be difficult to assess, where the imagery is only available from one perspective

In addition to the above, a desk top approach will not provide information on:

• The structural integrity of the roof and consequentially the need for additional strengthening

• The ease of locating the inverter in the roof space, it must be as close as possible to the panels to avoid energy loss from the DC current produced by the panels

• The costs of the wiring from the inverter, typically in the roof space, to the consumer unit. While external routing of the wiring is fairly easy to quantify and cost, internal routing its feasibility and costing is more of an issue.

• The condition of the roof and hence the potential need to undertake roof repairs prior to fitting the PV installation

In our experience, undertaking some site surveys is an essential next step in the feasibility process, which the social landlord can either undertake themselves or delegate to the installation company there being pros and cons to both approach. We would however always advocate the desktop study approach first, in that in narrows down the number of properties that need to be surveyed and ultimately they all will need to be surveyed, as part of the installation process. On site surveys can be undertaken for each property, or where being commissioned by the social landlord, more typically on an archetype basis. An on-site survey would involve:

• Confirmation of property orientation, identification of any shading issues and accurate calculation of the roof pitch.

• Accurate calculation of the roof size and hence PV system size.

• Inspection of the condition of the roof to ascertain any repairs required prior to installation of the PV system.

• Inspection of the roof space to determine the construction and condition of the roof structure and hence the need for any additional strengthening to deal with the additional roof loading.

• Design of the system Inverter location in the roof space and cabling back to the consumer unit.

This information can then be used to re-calculate the system costs and financial feasibility as described above, allowing the social landlord to validate its desktop conclusions or to modify its conclusions as to which properties are appropriate for PV installation.

In our experience assessing the structural integrity of the roof is often overlooked by installers. For a recent feasibility that we conducted for a North East local authority a third of the properties that we assessed required additional strengthening as assessed by our structural engineer. While none of the strengthening costs were significant, failure to identify and manage the risk could have given rise to a significant cost burden to the local authority in question in later years.