Retrofit: developing a Decision Making Matrix

I’m on a train with time to carefully consider an email from my friend Dr Ben Croxford at University College London, which has prompted this posting on our research into decision making matrices for retrofit:

One of our objectives in our two Technology Strategy Board ‘ Retrofit for the Future’ projects is to develop a decision-making matrix based on evidence of cost and benefit, in relation to project funding available.

A well-researched paper was presented at the international passivhaus conference a couple of years ago that concluded in terms of both economics and building physics, that it is silly to do less in a retrofit than a combination of insulation (external insulation including walls, foundations and roof and ground) combined with high performance windows. To do one without the other risks condensation problems and unnecessary additional costs resulting from the need for future upgrades. After this, other things like a heat recovery ventilation unit (which is recommended for health), an efficient boiler, efficient lights and appliances are relatively cheap and easy to do but if really necessary could be delayed.

This fabric first approach is often thought to be too much effort by people who think it is enough to tinker with a boiler, lights and perhaps even a few water butts. The evidence I have seen to date indicates that to get 80% of the possible energy-saving benefits, there is no escaping the need for insulation and well-fitting windows. The question for the most worthwhile and interesting investigation and cost-benefit analysis is how much insulation? For example, either:

(a) 100mm external insulation and don’t alter the eaves and gutters, or

(b) 250-300mm external insulation and extend the eaves and gutters and improve the cold bridging performance at this junction at the same time?
And another issue worthwhile investigating is:

(c) Triple glazed windows with insulated frames, or

(d) Double glazed windows
The answer to these question will be found in some kind of three dimensional matrices, I think. There will be a different matrix for each house type and for the first question above, the factors to consider are: thickness of insulation and cost; impact of insulation thickness on energy consumption and cost saving; technical complicating factors (such as eaves overhang, condensation risk, planning restrictions), and financial complicating factors (such as total pot of money available, feasibility and relative cost of phasing insulation layers, fuel costs & bank interest rates). A matrix or a group of matrices that consider(s) all these factors could potentially provide objective advice on whether it is best to spread an available pot of money over a few houses or spread the money relatively thinly over a larger number of houses.

Such a matrix must be based either on a large number of built examples (but funding is unlikely to come quick enough) or alternatively on proven methods of design and real-life costs. We are currently working with Universities to test the accuracy of our design predictions using PHPP software and analysing actual tendered and built costs in detail with the help of a cost consultant (Richard Whidborne) and a research student. My aim is that within a year we will have sufficient data outputs from Technology Strategy Board funded University research to tell us whether our design tools are sufficiently accurate to be able to make worthwhile predictions of energy consumption in relation to different insulation thicknesses. In the meantime we are assuming that our meticulous energy-in-use calculations are going to be proven to be accurate enough for the kind of matrices I have outlined above, and we are also working on detailed costs in order to develop matrices that provide answers to the questions outlined above.

If any more students are interested in helping with this research, we would be glad to hear from you, email

Justin Bere

To reply to a Comment: Click the “Reply” at the bottom of the comment.