Notes from the (thermal) edge: Part L Compliance (2 of 2)


The following post was submitted by Joseph Little, Architect and environmental consultant, on 13th October 2014. If you are unsure as to terminology used please see Introduction to thermal bridging: The basics of thermal bridge heat loss (courtesy of Joseph Little Architects).

IF you are a Certifier and you don’t know what the following conversation means, or if you don’t know what a temperature factor (fRsi) is, you probably should not be certifying Part L compliance.


Author: Joseph Little, Architect and environmental consultant

Some aspects of compliance with Part L of the regulations are far more complex and costly than they need to be. Currently this complexity isn’t fully recognised in the Department of Environment or by Industry. Accredited Construction Details (ACD’s) are sets of junction details for various types of construction that have been proven to have relatively low thermal bridges and no risk of surface condensation: the related psi-values for selected junctions are used to generate a Y-value (a figure that represents the full amount of non-uniform building fabric heat loss as an expression of the area of the building envelope) in the Dwelling Energy Assessment Programme software (DEAP). After many other inputs the latter generates a building energy rating (BER), an energy value, a CO2 emissions indicator and an energy performance coefficient (EPC) for a dwelling. Several outputs from DEAP, including the EPC, are used to demonstrate compliance with Part L (bearing in mind that other aspects of Part L compliance, such as thermal bridging, are proven without use of DEAP).

There are two default Y-values: 0.15 W/m2K which may be used for all dwellings of any age where the ACDs don’t apply, and 0.08 W/m2K if all details comply with the ACDs. Alternatively one can calculate a Y-value using the ACDs (which can give a value better or worse than the default of 0.08 W/m2K depending on geometry) or a mix of psi-values from the ACDs, manufacturers catalogues and bespoke thermally modelled details, etc.. If a designer is able to use ACD details appropriately and consistently for the design of a particular dwelling s/he will be able to calculate a Y-value easily (check out the SEAI Y-value calculator), simplifying one aspect of creating a BER and EPC.

So far so good.


The problems are that:

(a) The thermal bridge heat loss of ACDs details (which were published a few months after TGD L(2008) and are closely based on even older UK details) is not sufficiently low to result in the kind of Y-values one should aim for (i.e. 0.04 or 0.05 W/m2K) to achieve a compliant maximum permitted EPC in DEAP without resource to unnecessary amounts of renewables. Indeed some ACD junctions such as external wall insulation – window junctions are badly detailed and result in relatively high thermal bridges.

TBs typical house_BLC copyright 2014(b) The ACDs are incomplete: this means, strictly speaking, that NO Y-value calculation can be made for even the simplest housing estate dwelling without thermal modelling. In the ‘zero tolerance’ world of SI.( (2014) that matters. For instance, the ACDs don’t contain a door threshold detail of any kind (though all houses have entrances), nor a detail for patio or folding door thresholds (that could constitute a significant proportion of the ground junctions in modern houses), nor a detail for a floor over a garage, a floor or roof of a window projection, a warm slab or raft slab, nor any sort of dormer or velux type window, nor even an EWI sill (excluding the bizarre Appendix 2 detail);

(c) The Department of the Environment recently confirmed that changes to the ACD sheets have been completed but not yet implemented. The changes are already shown in Appendix D of TGD L(2011). The following is a quote from an email I recently received from the Architecture/Building Standards Section “L 2011 encourages the development of customised details by certified designers.  This is a conscious effort to spur the market towards informed project specific solutions rather than generic approaches.  We have no plans to produce higher performing ACDs for generic use”. Hardly a case of the Department trying to support small, hard-pressed practices. Therefore Architects that want a wider range of lower thermal bridging details must either wait for manufacturers of major construction systems to calculate exactly the detail needed, or draw it themselves and pay for its calculation by an independent thermal modeller.  So far architects are staying away from thermal modellers in the NSAI scheme ‘in droves’: the market has not been activated.

(d) Default psi-values (that are deliberately high, i.e. poor), such as have been used for some years in the UK, have been proposed by SEAI but not yet adopted. It is not clear why. These values can be a useful way of including short, or seldom used, non-ACD junctions in a Y-value calculation. This encourages designers to undertake Y-value calculations and spend their limited money on calculating the longest or most improved details. Over time they will naturally calculate more details improving their Y-values further. As with any calculated Y-value the length of the junctions is measured, adding significantly more accuracy than a default Y-value.

(e) There is no methodology (such as exists in the UK) that guides an architect or BER assessor through the selection and calculation process of creating a Y-value. For instance the UK methodology states that roof or wall junctions that subtend an angle of less than X degrees can be ignored, psi-values for Y detail type can be substituted for Z type etc. These kind of rules allow a sane and relatively simple way of calculating a sufficiently accurate Y-value. Ironically they encourage low thermal strategies and encourage the use of thermal modellers. Importantly this methodology also allows proof of compliance with procedure: architects, and even more so certifiers, need confidence and a clear compliance path. To date SEAI, NSAI and DECLG have each seen the question of whether a methodology should be created, or simply adopted, as an issue for a different Govt. Department: it is my view the responsibility rests with the Department that ‘owns’ DEAP: SEAI.

In summary without default psi-values and a methodology, in a context where a simple housing estate house may have between 18 and 25 junctions and a unique design could have more than 50 architects face a costly, complex and uncertain path in addressing the thermal bridge impact of their designs. Ironically, in the context of a, b, c, d, e above, installing more high maintenance ‘solar bling’ on the roof is currently an easier way to get an EPC-uplift: it’s also easier to prove its compliance.


Part L demonstrates two key areas that undermine the absolute guarantees implicit in SI.9:

  • The actual technical guidance documents themselves are flawed and contradictory (in this case incomplete).
  • Compliance requires expertise that currently is not readily available in the industry.

The hands-off approach by the government, “here’s the regulation now sort it out” has put practitioners in the unenviable position of having to issue absolute guarantees of compliance for situations that are at best unclear, and at worst non-compliant.

Other posts of interest:

Part L compliance issues – S.I.9 (1 of 2)

Part L- is compliance worth the paper its written on?

Design Certifiers – 3 things about certifying Part L… 

Why the design certifier and architect need third party building fabric assessments

Opinion piece: new building regulations and materials risk analysis

SI.9 and Part L | Specialist ancillary certifiers Part 2

SI.9 and Part L | Are specialist ancillary certifiers needed? Part 1 

Dispensations and Transition Arrangements

Practical Post 10: No retrospective compliance – BC(A)R SI.9

Practical Post 13: Duties & conflicts- BC(A)R SI.9 

1 thought on “Notes from the (thermal) edge: Part L Compliance (2 of 2)

  1. d7h

    Hi there,
    apologies for coming to this post so late.
    notwithstanding the huge difficulties caused by the incomplete and inadequate acd’s which we are effectively forced to use to prove part l compliance in our black and white world of opinions… what also are the consequences for us as a profession of using these department of the environment details for the buildings we propose to build?… i’ll never have to draw a detail again if i want to be able to sign off on compliance later on and as such it seems i’ll never want to design a building that might have a detail not on the ‘approved menu’.
    deap is also black and white on this and requires ALL junctions to be compliant and so one special junction that was maybe not considered or which can’t be lifted off the doe acd menu will bring the 0.15 figure into play, and it is almost impossible, and cost-prohibitive, to be part l compliant if 0.08 isn’t used for the thermal bridging junction figure in deap.
    i can’t believe this post has no comments…sigh.


Leave a Reply

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