Previous post:

Next post:

Part L1A, SAP 2009 and the fabric first approach

by Mel Starrs on February 21, 2011

in Part L

I’m loving my job at the minute. The office has a huge volume of flats currently at roughly RIBA stage D for a PFI project we are looking at. I’m doing initial SAP work to inform the bid and it has given me the chance to poke about under the bonnet of SAP 2009, Part L1A 2010 and CSH 2010. The volume of work gives me the chance to do a little more comparative work than I would normally get the chance at. This informs the bid and is a little more interesting than grinding out compliance calculations.

What follows is as much an aide memoire for myself, as information for others. Part L continues to get more and more complex, which the engineer in me loves, but the realist in me fears. If you are a SAP assessor reading this and you’ve never tried to fill in the SAP spreadsheet by hand, looking at each parameter as you go, have a go. It’s very enlightening and far more use in consultancy than sitting in a training course for the software (sorry to software training providers).

One of the main changes to CSH 2010 has been the addition of credit ENE2 Fabric Energy Efficiency (aka FEE).

This concept of fabric energy efficiency was developed by the Zero Carbon Hub back in November 2009. It is strongly anticipated that it will form part of the definition of zero carbon when that is released.

So if you are looking to develop a future-proofed, fabric first approach to building houses, meeting the FEE targets for levels 5&6 of the Code today would make sense (as they are likely to be the minimum standard for Part L 2016, possibly with an interim measure in Part L 2013). And in theory, if you build to this fabric standard now, you may get the chance to convert your development to zero carbon in the future (let’s not debate the logic of this in this post).

So what is the definition of FEE then? What does it mean? The illustration belows trys to explain:

It’s a little bit complicated – the fabric performance of the building, plus some (but not all) building services (but not necessarily the one’s you intend to specify). Hot water gains and MVHR effects are ignored. See details below.

FEE in CSH is a replacement of HLP (heat loss parameter). ZCH did investigate retaining HLP but ultimately it was rejected as it is not used in other countries and although it includes heat losses via fabric and ventilation, it does not include passive solar gains or internal gains. (All this talk of passive and internal gains starts to point towards Passivhaus – but I’ll not go into that in this post).

Let’s go back to basics for a little while. Back in 1997 when I started at Max Fordham’s the first equation I learned was the very simple heat loss equation*:



Heat loss Q=kW

U value =W/m²K

Area =m²

N=number of air changes per hour

Volume =m³

Delta{t}=difference between internal and external temperature

The first part of the equation deals with fabric heat loss and the second the ventilation loss. These were simpler days, before carbon compliance. All we needed to worry about was the boiler big enough to heat the radiators on a cold day. An ideal room temperature of say 21°C with an external temperature of -4°C, with an internal air change rate of 1 for a leaky older building, work out the fabric heat loss from the U-values and add a factor or two and we were done. To make life easier, we used Excel on a computer to spreadsheet up the dimensions and speed up calculations (prior to this everything was done by hand, kids). I’m being slightly flippant, but I want to illustrate the basics of where some of SAP comes from.

Back to SAP. If you refer to the first few sections of the SAP Worksheet v9.90 it starts to look quite familiar to the above equation. Section 1 covers physical dimensions of the building. Section 2 covers ventilation (in more detail than our good old rule of thumb – 1 for a leaky building, 0.25-0.5 for a new build). Of course, since 2006 we have been building to air leakage targets. Section 3 looks at Heat losses and heat loss parameter. Again, this looks familiar. U-values and areas. And at the end of the section is thermal bridging. Thermal bridges that occur at junctions between building elements are included in the calculation of transmission heat losses.

The quantity which describes the heat loss associated with a thermal bridge is its linear thermal transmittance, Psi. This is a property of a thermal bridge and is the rate of heat flow per degree per unit length of the bridge, that is not accounted for in the U-values of the plane building elements containing the thermal bridge. To calculate Psi values I would recommend using THERM and referring to Peter Warm’s excellent resources available here. Then a copy of BR 497 Conventions for Calculating Linear Thermal Transmittance and Temperature Factors (BRE Reports) a few evenings reading and tinkering, and you’ve saved yourself £2100. It reamins to be seen what qualifications BCB will require from those submitting calculations under route 2) below. Am I a person with “suitable expertise and experience” or do I need a piece of paper to tell the world?

Refer to page 80 in SAP 2009 for Appendix K: Thermal Bridging. Thermal bridging was previously contained within SAP 2005, but there was little incentive to calculate in detail:

The value y = 0.08 applies for new dwellings whose details conform with Accredited Construction Details (or are otherwise shown to be equivalent), and a default value of y = 0.15 applies otherwise.

In 2009, we now have to calculate the y-value:

There are three possibilities for specifying the thermal bridging:

1) All detailing conforms with Accredited Construction Details or another government-approved source involving independent assessment of the construction method. In this case – use values from the ‘accredited’ column of Table K1 for Accredited Construction Details, or – use the values provided by the approved source in equation (K1) along with the length of each junction. Here ‘Accredited Construction Details’ means:

– For England & Wales and for Northern Ireland: Accredited Construction Details, as listed on

– For Scotland: Accredited Construction Details (Scotland) Environment/Building/Building-standards/publications/pubtech/techaccconstrdetails

2) Ψ values have been calculated by a person with suitable expertise and experience in accordance with BRE IP 1/06, Assessing the effects of thermal bridging at junctions and around openings, and BR 497, Conventions for calculating linear thermal transmittance and temperature factors, but have not been subject to independent assessment of the construction method. In this case the Ψ values are increased by 0.02 or 25% (whichever is the larger)9 and used in equation (K1) along with the length of each junction.

3) If neither of the above applies use y = 0.15 in equation (K2).

However, as method 1) was not fully in place by October 2010, the penalty to method 2) was temporarily lifted:

Interim basis from 1 October 2010 until this Appendix is revised following the introduction of Accreditation Schemes for construction details:
Values calculated using option 2) below are treated as accredited and the addition of 0.02 or 25% is not applied.

If pursuing the ACD route confirmation needs to be accompanied by the ACD Drawing Detail Reference Number for all junctions that will follow ACD, E.g. Suspended Ground Floor Slab/Detail Number-MC1-GF-01. It is no longer acceptable to provide a written confirmation without the Drawing Reference Number of the ACD details used.

ACD details can be found at (this is a different link to the one quoted in the SAP 2009 document, and the correct one)

Junctions that need to be confirmed:

External walls junctions with:-

  • Lintels, sills & jambs
  • Ground floors
  • Intermediate floor within a dwelling (for Houses)
  • Intermediate floor between a dwelling (for flats)
  • Balconies within a dwelling (for Houses)
  • Balconies between dwelling (for Flats)
  • Eaves (Insulation at ceiling level)
  • Eaves (Insulation at rafter level)
  • Gable (Insulation at ceiling level)
  • Gable (Insulation at rafter level)
  • Party walls between the dwellings
  • Flat roof with/without parapet

Party walls junctions with:-

  • Ground floors
  • Intermediate floor within a dwelling (for Houses)
  • Intermediate floor between a dwelling (for flats)
  • Roof (Insulation at ceiling level)
  • Roof (Insulation at rafter level)

While this evidence may not be asked by the BCB for Design Stage submissions, it will definitely be required by BCB for the As-Built Stage submissions.

For those familiar with Passivhaus, you will see a big difference between PHPP and SAP here – the difference between using internal and external dimensions has quite an effect on how thermal bridging is approached. What I’ve not tried to do yet is run a PHPP building (thermal bridge free) through SAP 2009. Anyone else done this yet? For those not familiar with PHPP, you just need to know that to compare the two without appreciating the subtleties between the two calculation methods can be a brain taxing exercise.

HLP is still calculated within SAP 2009 (SAP spreadsheet – box 40). HLP is measured in W/m²K and is a measure of the impact of both external surface area, insulation value of construction and airtightness. The previous version of CSH rewarded a lower value for Heat Loss Parameter. It encouraged the design of efficient built form such as flats and terraces as well as increased levels of insulation and airtightness. Values of <1.1 gained 2 credits under ENE 2, <1.3 gained 1 credit. Generally, flats scored well under CSH in this respect and gain 2 credits.

What is the effect of looking at FEE rather than HLP? First off, what is FEE?

From SAP 2009 page 32:

Fabric Energy Efficiency is defined as the space heating and cooling requirements per square metre of floor area, obtained at worksheet (109) when calculated under the following conditions:

– natural ventilation with intermittent extract fans

– 2 extract fans for total floor area up to 70m², 3 for total floor area >70m² and up to 100m², 4 for total floor area > 100 m²

– for calculation of heat gains from the hot water system worksheet (46) to (61) inclusive and (63) are set to zero (equivalent to an instantaneous water heater)

– 100% low energy lights

– column (B) of Table 5 is used for internal gains in the heating calculation

– column (A) of Table 5 is used for internal gains in the cooling calculation

– overshading of windows not less than average (i.e. very little is changed to average)

– no heat gains from pumps or fans

– the heating system has responsiveness 1.0 and control type 2, no temperature adjustment, temperature and heating periods according to Table 9 irrespective of the actual heating system

– cooled fraction is 1.0

Other data items are those for the actual dwelling. The above are special conditions for calculation of Fabric Energy Efficiency and do not apply for SAP calculations.

So it’s a heat loss rate, but for a very specific set of criteria. Almost another notional building. The intention is to be able to compare different dwellings against each other on a like for like basis. I can see the logic but I worry it might be difficult to ball-park what a building might be at early design stages. I’m hoping it comes with experience.

So let’s look at some examples. I’ve run 9 sample flats in a block of 52. I’ve used good to best practice u-values and good practice air tightness (better than compliance, but not as stringent as Passivhaus). This set were run with ACD calculated from those published in the link above. Under SAP 2005 if using ACD, I could have made the assumption that y-value = 0.08. As you can see from the graph below, most of the flats still cluster around the 0.08 – 0.09 value. However, I have 2 outliers, both for mid-floor (MF), mid-terrace (MT) flats which are much much higher, and one of these is greater than 0.15. There are many fewer junctions in these flats. You can also see that the FEE values for these flats is exceeding the target of 39 kWh/m²/yr by quite a margin.

As expected, end terrace (ET) flats perform worse than mid-terrace. We’re more or less meeting the CSH ENE2 requirements for level 5&6 without too much strain. For blocks of flats there is a calculator tool which allows blocks to be aggregated, so the effect of outliers is mitigated.

Lastly I compared the old HLP measure to the new FEE:

As I would hope to see, a more or less straight line relationship.

What I’ve not yet investigated in any detail is the impact of Thermal Mass Parameter (TMP).

A key parameter within SAP 2009 that can influence cooling is known as the Thermal Mass Parameter (TMP). This is established by defining the materials used to construct the dwelling, in particular the internal elements and their finishes. Whilst thermal mass needs to be linked to a solar shading strategy and night time ventilation to be affective, modification of the TMP within cSAP provides a useful sensitivity check for its influence on the overall energy demand.

TMP is an input into the SAP 2009 worksheet (box 35).

There is a free tool available from Concrete Centre and Arup which will calculate Kappa values (effective thermal capacity kJ/m²K). Indicative values would be fine generally (for my examples above I was using 100 as this is a worst case scenario. On flats which were overheating, it may be worth calculating the actual TMP to see if this alleviates the overheating).

As you can see, a fabric first approach is now much more detailed than good old Q=({sum{}{}{UA}}+{{NV}/3})Delta{t}.

Hopefully, it all now is a bit clearer. The more examples I deal with, the more I learn. Any thoughts and comments out there?

*Wordpress geeks might want to know I’m using this plugin to display the equation using these conventions