Maintenance of Windows - 2




Simon Berry - Letchworth Garden City Heritage Foundation

Robin Kent - Tangram Technology Ltd. 

 Measuring the hidden cost

Last month we looked at the issue of maintenance in windows and concluded that there were significant costs associated with maintenance and that these could significantly affect the choices made under Best Value purchasing. This is particularly relevant when the experience of Local Authorities and the real costs of window maintenance are revealed. It is important to realise that the costs discussed are independent of frame material - frame maintenance is additional to the maintenance costs of the hardware and operational adjustments.

The attitude encouraged by Compulsory Competitive Tendering (CCT) was along the lines of 'We have £XXX pounds in the budget - lets get as many windows installed for this price as possible'. The change to Best Value (BV) will begin to encourage a change to: 'We have £XXX pounds in the budget - lets get as many good windows installed for this price as possible and minimise the total cost of the contract'.

A true assessment of the total cost may result in fewer windows being installed initially but the long-term costs will certainly be less. This is where CCT went wrong. The emphasis was on numbers and the initial purchase cost - not on real value in the long term. In many cases the initial purchase cost is nearly irrelevant in the long-term. Specifiers need to use models of the cost behaviour of windows over time. Martin Harrison developed the earliest model in his book 'uPVC windows in the 1990s' but this was based largely on frame maintenance and assumed that hardware was not replaced in the 30-year life of the window (also assumed). The Berry study shows that maintenance of hardware (both adjustment and replacement after failure) is a prime cost element that needs to be included in any model seeking to get reliable Best Value results. First order estimates from the Berry study show that at least £300 per dwelling needs to be allocated as a maintenance cost over the 30-year life cycle simply to cover hardware maintenance. In fact the long-term costs will almost certainly exceed those found by the original Berry study, as most of the windows installed were relatively 'young' in terms of a 30-year life cycle. Things can only get worse. We need to change the way we look at the cost of windows and the allocation between capital and expenditure budgets - it is nonsensical to engage in a capital investment programme based on low initial costs only to find that this causes excessive recurrent expenditure. CCT has already shown the error of this and BV could well go down the same route unless real Best Value is sought and maintenance is included in any cost model.

A first attempt at developing a cost model using these concepts provides some guidance on the likely costs for the future. The output of one set of assumptions is shown in the figure. As with any model the output depends on the initial values - we can only start to develop the model - what you put into it is your responsibility. The first sample model is available free of charge from Tangram Technology (Download). The model can be developed further - send us your suggestions for improvement and we will update the model. Collaboration in building a realistic model will help everybody involved.

The high costs revealed in the 1999 study by Berry (Berry - 1999) were based on well-recorded outsourced reactive maintenance. The cost of a single visit was around £67 but even this appears low if all the costs are included. Some estimates range as high as £150 per visit when all costs are taken into account, other participants in the study reported very low costs but these were achieved by carrying out the maintenance internally and ignoring other cost elements such as administration (it is only an overhead!) and the cost of parts. Every authority will establish their own costs but these must be real and reflect all the costs instead of only some of them.

The maintenance costs found by the Berry study were almost purely concerned with hardware maintenance, not one instance of PVC-U frame failure was found. The PVC-U frame material did not require maintenance within the period of the study because the windows were relatively young in terms of their life cycle. A frame maintenance cost will also need to be included to reflect the differing maintenance loads of the frame material.

It is going to get worse

The one certain outcome of this work is that the concern is going to get worse. Many large-scale Local Authority replacement programmes have not used a Best Value approach and the windows installed will suffer from increased maintenance loads in the future. Failure to plan for the inevitable window maintenance loads leads to unplanned future expenditure. If we assume that the 468 UK Local Authorities each have a housing stock of 10,000 dwellings then this expenditure could easily rise as high as £46 million per annum. This will inevitably get worse as hardware and frames approach the end of their service life and have higher maintenance loads. A new industry in maintaining windows may well be created to service the inevitable demand. As always, today's production is tomorrow's liability.

The Housing and Related Services Performance Indicator BVPI 65B, 'Average weekly costs per local authority dwelling spent on management and repairs' ( requires an authority to seek a reduction in the average overall maintenance costs per dwelling. Concentrating on the maintenance load of the frame material has hidden the very real concern of hardware maintenance. Action needs to be taken now to establish economic measures for reducing the overall maintenance costs for the future.


Window Maintenance

The Window Maintenance series is designed to raise awareness of a critical subject in the area of window technology. It is being published in Fenestra Journal on a monthly basis and published here after the Fenestra publication. The series is

Part 1: The Changing Landscape
Part 2: Measuring the hidden cost  (This section)
Part 3: Reducing the hidden cost - Strategies for the future

Last edited: 11/03/10

© Tangram Technology Ltd. 2001

Our standard disclaimer regarding Internet data applies.