Now I do just about remember how radiator heat outputs are specified. It’s all about Δt, the difference between the average temperature of the radiator surfaces and the temperature of the surrounding air…
For normal gas-fired boilers, it’s quoted for Δt=50C, so for an ambient temperature of 20C, that assumes a radiator of 70C. That’s 75C at the top and 65C at the bottom, plus the assumption that water balances itself out. Because some things have to be nailed down for system specifiers, manufacturers quote for a 10C difference across the radiator, and provide tools for deciding what size radiators, pump, boiler, etc. on that basis. They also sometimes give equivalent tables for some smaller Δt’s because of the increasing numbers of low temperature systems.
What St Columba’s has is 6 Stelrad Elite 600 x 2800 K2 radiators, plus one <mumble/>. These are just like in a standard Edinburgh domestic system – K2 means the double finned ones that put out the most heat for a given length because there’s more surface area for the air to contact. I can remember seeing claims once that dusting the fins on one of these could increase the heat output by 5%, by un-insulating the metal I guess. I never traced that back to a proper source, but I can just picture the PhD students as they work on that one!
Under the standard benchmarking conditions these radiators will be 5 kW each. For anything besides 75/65/20, the physics, and therefore the performance, is determined mathematically (tain’t the universe wonderful?) as explained and aided by the Engineering Toolbox. (Standard internet caution applies here; equations and calculations that truly matter are best checked with verified sources. I have no idea who maintains this page.) Still, once the radiators are unblocked this kind of graph, along with air temperature, should bear some resemblance to the data sheets. Of course, that doesn’t tell us anything about the effects of blocking up the airflow.
I still think it’s worth getting the data, and here’s why. The drop across the radiator depends on the flow rate. Faster flows mean a smaller drop and a higher average temperature; slower increases the drop and lowers the average temperature. Sometimes that’s a useful thing to do, depending on the size of the boiler compared to the radiators. If the boiler is over-sized and the radiators are under-sized, as seems to often be the case in churches, then it might be possible to increase the heat output by turning the boiler thermostat up and slowing the flow. In addition, if condensing boilers really only condense when the return is under 55C – something we still need to verify! – then there’s an extra bonus to ensuring the system is set up that way.
If I remember correctly (which I never do!) St Columba’s has a variable speed pump, but not a condenser. I’m not convinced we’ll learn anything useful for this site from this part of the measurement (as opposed to the gas consumption), but I can almost guarantee working through this exercise will help us understand some of our other sites and let us perform a personal mercy mission. It’s not a church, but a quick test on my living room gave 46/30/15. It’s no wonder we only use it for storage!
The old graph does raise an immediate question for me – is the green line for real? If so, then there’s a balancing problem; like maybe the lockshield valve is closed too far. It seems unlikely, though. It’s more likely a problem with the thermocouple’s contact to the radiator. I don’t actually know how far off the readings are because we’re mating thermocouples with cylindrical housings (waterproof DS18B20s packaged just like these) to flat surfaces. I’m minded to try a pre-test at home – one thermocouple with plain plastic tape and another at the same level insulated on one side by some insulator – thermal curtain lining or bubble wrap, or if they mind being 75C, maybe a piece of old oven glove.