St. Columba’s by the Castle – where we ran an Innovative Learning Week event last year – has an interesting follow-up question for us. How much will it help the space warm up when they remove that wood paneling around the radiators?
Whenever I look at them, I think “this is a place that takes the name ‘radiator’ seriously” – after all, there will be radiation off that nice exposed surface. Unfortunately, radiators are misnamed. They really heat spaces almost entirely by convection. These are designed to take cold air in at the bottom, heating it so it rises and comes out the top. You can’t see from the picture, but there is actually a grill in the shelf above the radiator. Not so underneath. The air intake is seriously obstructed, whereas the air outflow is just slowed down by extra turbulence, dumping heat right there instead of further into the space.
Answering their question starting from first principles is tricky. There are data sheets that will tell us how much heat output this radiator would give if unobstructed, but of course, not what happens with all of the interesting things that happen to them in practice. And I know from talking to Dimitri about building a test rig for the hot wire anemometry that turbulence is tricky stuff! They’re really looking for “how much more heat output will we get, and will it be enough or do we need to make more infrastructure changes”. It’s a bit hard to mount equipment on the radiator for the “before” picture. These are my thoughts about how we might answer the question.
The first thought is: refine the question. They’ve already asked a consultant to verify our suggestion that the heat output was getting choked, and have decided that the panelling is coming off, at least to the degree that it should allow the radiators to operate efficiently. The consultant didn’t suggest adding more radiators or other heat output in the space, although they did suggest a warm air curtain above one of the internal doors to reduce heat loss during use. Even if people like the panelling, we don’t need measurement to finesse exactly how much needs to come off; we just need to follow the manufacturer’s recommendations for clearance. Ironically, that might mean that the grill in the shelf goes – once there’s proper airflow, sometimes people add shelves above radiators just to divert the airflow into the room and reduce drafts! I think the question is really motivated by two real issues, though: a genuine anxiety about whether it helps enough or whether they should go straight on to doing more, and a scientific curiosity that turns out to be pretty characteristic of this congregation. So we really do want to provide before and after assessments, if we can.
My second thought is: this is almost the kind of application for which a duct anemometer is designed. Turn on the system, shove the anemometer above the radiator to measure the air speed and temperature differential across the radiator; work backwards from the radiator. data sheet to figure out what percentage efficiency we have in practice. For extra credit, take anemometer readings above the grill and in the gap above the radiator, so we can estimate how much of the loss is at the top. I think we’ve decided previously that we can’t get anemometry onto the radiator to do this (gaps not big enough/no access hatches), so our before measurements have to be what we previously got, but we should revisit that in this very last chance before the panelling actually comes off.
What congregations care about most is getting the place comfortable despite only occasional heating. With the panelling off, all other things being equal, the air temperature should rise faster. They’re never equal, though – external temperature, wind, how damp the walls are, the amount of sun in the previous three months – these things all have large effects. So we can eyeball the air temperature against a range of previous warmup graphs, but even if we happened to have a good match in our past collection, we wouldn’t necessarily know it was. More to the point is the heating system performance. What we want to know is – did taking the panelling off mean the radiators were able to operate more effectively? In the absence of before-and-after anemometer and surface temperatures, that means using a surrogate: gas consumption. If they can dump heat better, the boiler will run more. We have one previous set of pictures trained on the gas meter, and this time we’ll be taking more. The weather conditions are different, of course, but I think at least there’s a known relationship between air temperature and how fast radiators can dump heat, allowing us to factor that out. It’s time to read the data sheets again to make sure we understand what matters.
In a space like this, heat losses are high so it’s hard for the building to reach steady state. I think we’ll want to divide gas consumption into two phases:
(1) warm-up, considering pipe/radiator temperatures and gas consumption up until the radiators reach the maximum temperature differential.
(2) the heat dump, when the gas consumption slows because there’s warm water coming back on the return, the air temperature rises, and maybe the differential across a radiator slowly drops because it can’t transfer heat to warm air as fast.
I guess I’d expect with the panelling off, compared to before, that the gas consumption and temperature differential for (2) will be higher. How much higher gives us the proportion due to the efficiency gain of unblocking the radiators, other things being equal – but since the rest of the equipment and settings are all the same, the confounds aren’t as bad as for air temperature. I’m not sure I know what happens during (1). Now I go back what measurements Iain and the university have made before, and Iain’s draft plan for when the panelling comes off; discuss anything that confuses me; and prepare the equipment.