I’ve been spending some quality time with spreadsheets and have an update on the way building regs for housing treats CHP. It’s not as simple as I thought here, although the result is similar. The incorrect method I wrote about a few weeks back is still in use, but it’s different from what’s happening in SAP 2005. Here’s a breakdown:

SAP calculates the emissions benefit from CHP by determining the carbon intensity of the heat produced. Under SAP, the carbon intensity of heat from CHP is equal to the emissions from the combustion of the fuel minus the avoided emissions from electricity generated at a central plant.  

In other words, in SAP all emissions from CHP fuel input are associated with heat before subtracting the emissions avoided at central power stations through the local generation of electricity. So if you look at blank 107* in section 12b of SAP you’ll see the formula for carbon intensity of CHP heat is, in effect, doing this:

((Ee / ηe) - Ee) / (ηh / ηe)
((Eg / ηe) - Ee) / (ηh / ηe)

Where:

Eg is the emissions factor for gas (0.194 kgCO2/kWh)
Ee is the emissions factor for electricity offset (0.568 kgCO2/kWh for SAP)
ηe is the electrical efficiency of the CHP engine
ηh is the thermal efficiency of the CHP engine

Conceptually, this is pretty convoluted. Another way to get to the same result is this:

 Eg / ηh - Ee x (ηe / ηh)

This sticks more closely to the way I described it above: first associate all emissions from fuel input to gas (Eg / ηh) and then subtract the emissions you avoided at a central power station because you generated electricity on site  [- Ee x (ηe / ηh)]. But whatever, the result is the same: electricity from CHP is assumed to have a carbon intensity equal to the grid electricity it’s offsetting and all the carbon benefit is stacked on the heat side.

Assuming Ee = 0.568 and Eg = 0.194, for an engine with an electrical efficiency of 30% and a thermal efficiency of 50%, the carbon intensity of CHP heat is 0.0472 kgCO2/kWh. Pretty low right? Except that since we stacked all the carbon benefit on the heat side, the carbon intensity of electricity from CHP is now 0.568.

So what sort of carbon savings does this give us? If we return to the same scenario I used in the last post on this topic it gives a good idea:

The baseline (using gas boilers and grid electricity) looks like this:

   Primary energy   Assumed efficiency   Useful energy   CO2 kg/kWh   kgCO2/yr 
Space heating    56.8 88% 50 0.194 11.0
Water heating    28.4 88% 25 0.194 5.5
Electricity    50.0 100% 50 0.422 21.1
  Total kgCO2 37.6

The CHP assumptions look like this:

Proportion of heat supplied by CHP     60% 
Split (CHPe / CHPt)  0.6 
Total CHP efficiency  80% 
Distribution losses  10% 
Backup boiler efficiency  88% 

And the CHP calculation looks like this:

  CHP Backup systems
  Useful energy demand (kWh) Demand met (kWh) Primary energy CO2 kg/kWh kgCO2/yr Demand met (kWh) Primary energy CO2 kg/kWh kgCO2/yr
Heat 75 45 50 0.0472 2.4 30 37.9 0.194 7.3
Electricity 50 27 27 0.568 15.3 23 23 0.422 9.7
Subtotal CHP CO2 17.7   Subtotal backup CO2 17.1
  Total kgCO2 35.75
Savings 7.7%

A savings of 7.7% looks pretty reasonable - in fact the results are exactly the same as the “common sense” approach in table 2 here, except that the distribution losses are handled slightly differently. A couple of things to note: for CHP the primary energy includes distribution losses but not efficiency losses. That’s because the CHP efficiency is built into the emissions factor for CHP heat. On the backup side, we go back to using 0.422 for grid electricity since we want SAP average rather than marginal grid carbon intensity.

So if we’re getting conservative figures from this method then how can SAP give you reductions of over 40% from CHP? Because SAP deals mainly with heat: on the electricity side, only lighting, pumps, and fans are included (and for a nat vent house served by CHP, SAP assumes no energy consumption for pumps and fans). It doesn’t include appliances, at least not yet.

Because all the carbon benefit is stacked on the heat side, the drop in emissions is magnified. Sure, under SAP all the electricity incurs a carbon intensity of 0.568 (there’s no differentiation between electricity from CHP and electricity from the grid), but there’s so little electricity included that this doesn’t matter. Here’s an example:

For a new three bed semi of 100m2, you might get energy consumption figures like these:

  • Space heating: 50kWh/m2
  • Water heating: 32kWh/m2
  • Lighting, pumps, and fans: 9kWh (assuming no energy for pumps and fans)

Plugging these proportions into the example above and the savings jump from 7.7% to almost 30%. Now increase the proportion of heat supplied by CHP to 90% and the savings jumps from 14% (using the original proportions) to 50%! SAP doesn’t mind where the CHP electricity goes - all the carbon benefit stays with the heat. Someone else incurs the 0.568 kgCO2 cost for each of those kWh’s from the CHP, but SAP doesn’t care about that.

In my opinion, the method SAP uses for calculating carbon emissions from CHP is a good one for a whole development since on this scale you can account for all of the electricity as well as heat from the CHP engine. But on a dwelling by dwelling basis, it’s misleading. Unrealistically, it takes advantage of all the carbon benefit without considering who pays the penalty. This will continue to be the case at least until SAP is revised to include all consumption within the dwelling.

Looking back on my previous entry I should have spent more time with the worksheets before lumping SAP in with other bad methods. I apologise Brian. In SAP’s case the method is good but the results are misleading.

One point to note: the discussion above underlines the big jump that will have to occur between Code 4 and Code 6 - not just because we’re going from 44% reduction in regulated emissions to net zero carbon, but because at Code 6 all emissions are considered. And as we’ve seen above, that may make things pretty difficult for gas CHP.