Thermal modelling and Psi value calculation

Thermographic image of wall

Heat losses due to thermal bridging at junctions have become increasingly important.

This is because more stringent government legislation and energy awareness is leading to increased insulation levels in the fabric of buildings. As of 1st October 2010 the Approved Documents L1A and L2A required that additional heat loss from thermal bridges be taken into account, with entry of detail lengths and Psi values for junctions being required to be entered into SAP and SBEM calculations (unless a worst case default is instead used). Previously a simple global y-value was allowed if Accredited Construction Details (ACD’s) had been followed – which is what most builders did.

The heat loss associated with these thermal bridges is expressed as a linear thermal transmittance, Ψ . The temperature factor ‘ƒ’ should also be calculated to determine the risk of condensation or mould growth, which can have significant health implications. Where construction details do not conform to generic Accredited Construction Details, they need to be evaluated using validated thermal simulation software and to follow agreed conventions and standards.  Modelling of the additional heat loss at junctions in the building fabric should be carried out by ‘a person with suitable expertise and experience’ following the guidance set out in BR497 and BR IP 1/06.

Currently there is no standard certification of modellers, no central depositary of details and the onus is (and will likely remain) on the thermal modeller to be able to prove competency.

However looking ahead, there is scope for a qualification and an accreditation scheme for the thermal modeller and for the details he or she produces. Previous discussions have been had regarding how Accredited Thermal Modelling schemes might work, but it seemed a huge undertaking and there was considerable scepticism as to how it could work, who would check design and as-built and who would pay for it all.

Some of the current roadblocks in my opinion are:

  • No-one is checking that given details have actually been followed and built (including ACD’s).
  • No-one seems to sign off or check process sequences have been completed and followed, or that junctions have actually been built as per the details.
  • There aren’t actually ACD’s available for a lot of junctions.
  • The ACD details don’t actually achieve the psi values attributed to them for all construction types and particularly for all construction U-values.
  • It’s a time consuming process and if modelled results come out worse than ACD’s (or worse than default values even), builders are likely to just claim the ACD’s Psi values.
  • There is no easy way to check as-built performance of junctions match claimed psi values.
  • There are so many combinations and potential details to model, that it gets a bit silly (and costly) requiring a fresh detail, process sequence and psi value for every possibility. Some common sense needs to be allowed.
  • If someone spends a lot of time and effort producing their own details, what stops other people using their work, but substituting with a comparable product? If that happens, who would want to commit to the expense, knowing your details will be copied?
  • There is no accreditation scheme for any of this at present and a business case to make a scheme work needs considerable thought.
  • There is no-one to check that modelled details have been modelled correctly and achieve the Psi values claimed.
  • There isn’t a qualification at present (although BRE were looking at this, but seemingly no impetus currently to rush to do so).
  • People already doing this kind of work (and there aren’t currently a large number) would probably be a bit annoyed if they have to pay someone to agree they are competent via a new qualification.
  • Some assessors are still just entering a y-value into SAP assessments, which then isn’t challenged by Building Control or picked up at EPC audit by OCDEA accreditation schemes.

This subject is one that needs further consideration moving forward, as the fabric heat loss becomes more and more important in the drive towards low carbon housing.

To keep up to date with all our latest blog posts you can follow the blog by clicking this RSS Feed link or by following us on Twitter @KingspanIns_UK or on Linkedin.

Share this blog post with your friends and colleagues by clicking on the social media icons below.

Print this page

Jon Ducker is a qualified energy assessor working for Kingspan Insulation Ltd. He has an extensive knowledge of energy efficiency, renewable energy systems and sustainability in buildings with an expert knowledge of the relevant sections of buildings regulations and standards and their interactions with SAP. He provides authoritative advice regarding energy assessments for a wide range of public and private sector clients.

0 comments on “Thermal modelling and Psi value calculation
  1. Agree entirely. I am building a new house to very high insulation standards, 200mm cavity and footings from strip concrete to finished floor have 200mm of polystyrene in the cavity and the inside face of the inner blockwork. However I get no credit for this because it is non standard. instead we have to add PV to meet a sap target.

    Is there way to calculate the Psi per linear metre degree Kelvin. (I have a degree in mathematical physics.) I know it can’t be used in the standard SAP model but if it makes much difference I will kick up a fuss.

    • You could get it modelled by an appropriately competent assessor in accordance with conventions in BRE 497 Conventions for Calculating Linear Thermal Transmittance and Temperature Factors and using appropriate software (TRISCO / BISCO; HEAT 2 / 3; STRAND 7 or various others available).

      For a quick sense check as to how much of a difference that might make and whether it’s worthwhile, try in your SAP assessment using a notional Psi value of zero with the actual length of junction (assuming that you haven’t gone for SAP default y=0.15) and see what the effect is on your TER/DER and TFEE/DFEE; or, (again as a sense check only), try a ‘y-value’ of 0.04 (representing a reasonably good level of thermal bridging for the dwelling as a whole) and see if you can get away with no PV. If you were close to passing in the first place, something other than PV may be a better bet.

      There is very much a lack of modelled details available and this needs improving upon, but it can be a time consuming process to model a junction and there are a lot of junctions and a huge number of possible ways that a junction can go together.

      Even if you were to get a junction assessed, the result can sometimes be worse than SAP defaults or the 2007 ACD’s might suggest. The defaults in my opinion are too lenient and the ACD’s need re-doing for 2013 and beyond.

  2. Hello,
    I have a client who is wanting to ascertain the thermal performance of insulation required below a wooden floor so they can establish the thickness of kingspan needed
    Is this something you could provide?

    Ian Brown
    SR Brown Ltd

    • Hi Ian,

      Thanks for your question. We suggest taking a look at our U-Value Calculator, which will help your client ascertain the amount of insulation needed for the floor to reach a specific U-value. Amongst other floor types, the calculator gives options for a timber floating floor, a timber floor on battens or a suspended timber floor. You will also need to work out the P/A ratio of the floor in question – guidance on this is given in this blog post. If you need further assistance, you can always get in touch with our Technical Department on 01544 387 382.

      Hope this helps.

Leave a Reply

Your email address will not be published. Required fields are marked *