Some simple comparisons

Last night, we thought we’d take a good look at the St Columba’s data and think about two things:  what do we know as a result, and what kind of advice could we give to people replicating our equipment and using it on their own buildings about how to know what the data means?

The major questions people have – like “would it be better if the heating system could bring the building up to temperature faster?”, “would we be better off heating more often?” and “is that a normal amount of gas?” – can be difficult to answer simply.  So much depends on context, and it’s not like googling about how to set up the system in a house or flat.  There just aren’t as many Victorian church buildings and halls around to be targeted with generic advice.

For Innovative Learning Week, Dimitri put together a resistance in series model of heat transfer in St Columba’s by the Castle, and we are learning from how that works.  (If “heat transfer” is the wrong term, blame the blogger – I was busy organising supplies when he explained it, so I’m going from what I see on the spreadsheet!)  But first, we’re taking a simpler approach of just looking at some comparisons among local buildings.  We think that if we get a set of the same kinds of graphs for a bunch of buildings, then we can start to get a picture of what they mean and how to tell whether something is out of line that will be more immediately appealing to those who’ve forgotten their physics.  It’s a first step, and something that could help with getting us started on an “In the Bleak Midwinter” campaign to start looking at more buildings.

So here’s the first simple comparison – St Columba’s and Christ Church Morningside, by the numbers.

                                                                                                  St C’s                                               CCM

Volume                                                                                1440 m3                                        3721 m3

Heat Output                                                                     33 kW                                              68 kW

Boiler rating                                                                     60 kW                                             200 kW

Boiler thermostat setting                                         6=82C?                                         5.5 = 77C

Max feed temperature reading                             70C                                                 72C

Time until return is fully warm                             50 min                                           90 min

Time to warm space*                                                  5.5 hours                                     8 hours?

Air temperature target                                              17C?                                                 16C

Gas consumption                                                          338 kWh                                        to be calculated

General heating pattern                                           Sun am, Thu eve                      Sun am, Thu eve

I hadn’t realized that St Columba’s is only 40% the size of Christ Church, but I never could pass the “fit the leftovers in the leftover container” test.  Time to warm the space varies tremendously with environmental conditions, and not just the current temperature.  Last summer’s cold and wet threw all of my usual models for how long to run the heating out the window – the building never got properly warm and took a bit extra until Christmas to make up the difference!  The St C’s estimate is from 20 Dec 15, with a starting external temperature of 9.5C, and I’m currently thinking about how to construct the best match from the data I have but 8 hours is broadly right.  Christ Church has lots of meter readings but not for 2015, so it’s a case of looking for broadly comparable conditions.

St Columba's - gas use versus air temperature, ILW event, Feb 2016
St Columba’s – gas use versus air temperature, ILW event, Feb 2016

As you can see from the numbers, the heat output in two churches match in proportion to their volumes – but Christ Church already knows from a consultancy report that more heat output would help.  In their case, they’ve been recommended to replace two radiators, each of which is roughly 3 kW, with 15 kW fan convectors – there’s enough boiler there to supply the needed heat.  So perhaps St Columba’s could also do with more output.  We need to think though – what size does the boiler need to be?  I’m not sure it’s a straight case of X kW of heat output requires X kW of supply – I kind of have the impression that the boiler is usually bigger and runs harder at the beginning to speed the process up. I’ve never really thought about it, though!

The one hint we have is Dimitri’s graph of the internal  temperature, external temperature, and gas use from a heating session at St Columba’s.  Blowing it up to cover just one heating session  tells us something.   He has the original labelled as boiler output rather than overall gas use, but I think that’s wrong – his graph over more time shows gas use at unexpected times, for instance.  But this gives us some idea of how much the boiler is running, especially if we think about what else was on.

St Columba's - observed data from Innovative Learning Week 2016
St Columba’s – a wider view

We’ll need to blog more examples from other sites – we have the technology, and in some cases we even have the data! – plus many more things.  I think the priorities, in line with questions we get asked, are how we determine the heating output for the various kinds of heating equipment (based on Martin’s stellar detective work at Christ Church where old cast iron radiators are concerned), plus our best guesses about how to know how much heat output is required.   We’re learning, anyway.  Over the summer, Jean may even tackle some remedial physics – and not before time.

One thought on “Some simple comparisons

Leave a Reply

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