Lou and I took HeatHack to the Informatics Jamboree, where we showed temperature measurement at height using balloons and microcontrollers, and qualitative air current assessment using balloons and fishing weights. It was very handy that the one place where it was safe to go to full height was just in front of what turned out to be pretty powerful ventilation outtake ducts! We actually had to walk the high level balloons over to the space after putting them up elsewhere, because nothing was going to go past that airflow fast enough to end up anywhere but in the lift.
We were too busy to take our own camera out, but the Informatics photographer has given us some of his happy snaps. It was all good fun and I found out a good few of my colleagues have pretty interesting backgrounds and skills.
Our experimental setup in the St Peters boiler room worked flawlessly – sort of. Every single one of the pictures is on the pi, uncorrupted. They just don’t show anything useful because the tape holding the pi up came undone. Oops!
We hadn’t intended to try it quite so soon, but after yesterday’s tour of St Peters, we realized we really want a bit of extra data – something (anything!) that tells us when the boilers actually go on. That’s because they have optimized start control, so they program it to be a particular temperature by a particular time, and the controller figures out when it should start based on the prevailing conditions. On the other hand, the building has a reputation for never reaching the temperature they’ve set. We can save a lot of time and head scratching if we can find out a bit about what’s actually happening in the boiler room.
We’re at it again – this time we have five Lascar loggers checking out St Peters Scottish Episcopal Church in Lutton Place. They’re taking the temperature every five minutes, and one is also checking the relative humidity. We’ve hidden them pretty well, we think, but still in reasonable enough places to provide useful data. We’re not encouraging an early Easter egg hunt – we’d like them to stay where they are! – but if you come across our equipment there, you at least now know what we’re doing.
We wanted everyone to have the experience of building the balloon clusters on Shrove Tuesday, which severely limited the measurement we could take, since we also didn’t want anyone to have to stick around for a complete heating cycle. On the other hand, we discovered that the techniques we invented for the day actually do work and have the potentially to be very informative. So we re-ran parts of the exercises on Thursday 12 March 2015. Here are the results.
In a late addition to our schedule, HeatHack has been invited to participate in the yearly celebration of all things computer science and beyond that is the Informatics Jamboree. It’s a party for the University of Edinburgh School of Informatics to let their hair down and show their families what they do – but we could all be brothers and sisters, right?
Friday 27 March 2015 at the Informatics Forum, 10 Crichton Street Edinburgh EH8 9AB
setup at 13:00
talk 16:15-16:45 – “How I learned to stop worrying and love Public Engagement: Beltane, I, und Du ”
Tuesday we did some prototyping for our Mini-Maker Faire display based on some cracking homework submissions. We discovered that mylar space blanket only looks conductive and that anything you can pierce easily with a through-hole LED also risks being fragile enough to rip. Contact at the tip of the LED leg is enough for a circuit, but hard to ensure on a rigid conductor. Steel strainers are conductive, but the holes are kind of small to thread shielded wire through. Bubble wrap makes a good, cheap light diffuser. And even though cut-up loo roll centres look a bit like tealights, there’s something else that looks even more like them: tealights. The outer frame slips off and is conductive to boot. This leads us to our design – tealight shells with LEDs poked through, balanced on a positively charged mesh, and with one leg dangling onto a negative charged length of copper tape mounted on foam to aid connectivity. Cute, yes? Continue reading Tuesday’s design result→
We knew before that the organ at Christ Church dumps cold air straight on the Michael’s head, but now we have the measurement to prove it. The hot wire anemometer showed draughts of .3-.6 m/s – that’s definitely in the “annoying draft” range and then some!
We also knew that the decorated organ pipes are just decoration – they just come out. What we hadn’t realized is that just behind the lower section, there’s a piece of plywood that’s been painted black. What’s less clear is why. It might be a modesty panel – so that the choir doesn’t get distracted by light reflecting off the things behind it when they should be concentrating on the sermon – or it might be a previous attempt to screen the organist from draughts. If so, it probably never worked very well. It’s bowed with age, and as Michael described it, “it’s basically a ski jump aimed straight at my head” – but even completely upright, the organist would have to be pretty short for it to help.
One way of making the conditions a bit better might be to reduce the air ingress in the organ chamber. The organ is under the tower, and it’s very windy even indoors up the top, but there’s a solid floor separating the two. Dimitri’s comments towards a formal analysis of heat loss are here:
My estimate is that the organ chamber contributes in the region of 0.5 m3/s of cool ventilation, with only the organist standing in the way of draughts in the region of 0.3-0.6 m/s, and significantly cooler than the surronding air (14 degree C).
And what happens to it afterwards? Well, I found the same amounts flowing at both ends of the choir pew at the back of the organist’s seat, so it seems to be adding up quite nicely. Even if the remarkably close match between the two figures was more down to luck than anything else, the matching orders of magnitude and the channeling of the air between the wall and the pews both make perfect sense. The draught is faster near the floor too (denser, cooler air). The dense cool air floods the floor of the church.
If it was the only source of ventilation, the church would be experiencing one change of air about every 80 minutes (putting in very approximate dimensions, will have to correct that).
Remember, it was a windy day! Dimitri has even provided a handy spreadsheet explaining his analysis. The church does know that wind strips the heat right out of the building, they just weren’t sure how. The question is, how do we stop it? Dimitri’s analysis estimates that the gap under the tower door contributes around half the total – .023 m3/s. Draughtproofing that door would be a good start. Then it might be time to look at secondary glazing for the organ chamber window. And here is where human factors come in: few people have ever even seen either. The tower has been off-limits for years – completely verboten by the owners of the previous mobile phone mast – and the organ chamber is a tight fit with the added spice of possible damage if you do the wrong thing.
Even if we can’t block all the air coming through, we might be able to change where it comes out so that it isn’t bothering the people in the space. That would make it so that the choir and congregation don’t have competing desires – at the moment the warmer the church is, the less comfortable the choir is. The simplest way might be to change the length and orientation of the ski jump, for instance. We’re talking about doing some experiments, now that we know about it. The materials need to be cheap and easy to modify, but not alter the acoustics. That’s very, very important – the organ is quiet and muffly already because of where it’s sited! That sounds to me like a good student project, if the community doesn’t get to it first.
It’s very important not to get the organ too wet or too dry – so much so that many organs have humidifiers, and some are in climate-controller chambers. Reducing the ventilation typically retains moisture, not sheds it. Organ tuners take regular humidity readings, but it’s easy enough to throw a sensor location up there off the same network as our readings in the centre of the nave. We just need to know when to doubt them. The sling psychrometer showed 48% at one end of the building while our cheap and cheerful DHT11 showed 39% mid-way down – worth testing more than once, but “close enough for jazz”, I think. We do know the building is drier now than during the wet winter. It appears to have been a low battery that sent the DHT readings sky-rocketing earlier in the year – we need to understand how that happens, so that we don’t worry anyone with incorrect readings.