Monday 30 April 2012

The logic of light switches

Here's an example from our bathroom. This is the wall inside the changing room.  The bathroom itself is beyond, through the door. As is typical of Japanese bathrooms, it is a wet room with a drained floor, so the whole room can be used as a shower rather than having to shower in the bath, worrying about splashing water on the carpet, as you often have to do in the UK.

The switch itself is on the left as you come into the changing room, then the door to the bathroom is beyond that, also on the left. One of the switches is for the light in the changing room, the other for the light in the bathroom. Which one is which?

Yes, I thought so too. I suppose when there are lights to the left and right and switches to the left and right, it's obvious that the switch on the right is for the light on the right. Like the ones below, custom-built for our low-voltage tension lights.

Actually those switches are vertically aligned too. They were going to be to the left and right though, corresponding to the two arrays of lights on the left and right of the room.

Anyway, when the switches are vertically aligned, it seems intuitive to me that the top switch is for further away, and the bottom switch is for nearer. So for these switches, the top one should be for the bathroom and the bottom one for the changing room. Don't you think so?

In fact, they are the other way around, and I still sometimes get them wrong. Maybe that's just me, or it's some fundamental difference between western and eastern visual grammar. Or maybe the switches were just connected at random.

Overproducing panels and overheating power conditioners

The display for the solar panels shows us how many minutes each day the electrical output is being limited by the power conditioners. There are two reasons it does this: one is that there is too much electricity coming from the solar panels, the other that the power conditioners are overheating. 

We have 9.12 kW of generating power on the roof, which means that the maximum theoretical generating power is 9.12 kW. In fact, the rating of each panel is 185 watts, but in fact their actual power output varies and all are above 185 watts, some up to 195. So in fact we have about 9.4 kW. Anyway, they told us we'd be fine with two 4 kW power conditioners. We believed them. Half of 9 is 4.5 and 4.5 is just a big 4, in the same what that 2 + 2 = 5 for large values of 2. Electronics comes across as this very exact precise science, but in fact it's as full of tolerances and variations as anything else. Carpentry can be a lot more precise. 

In the event, we've found that for a few hours per month the power conditioners are limiting the amount of electricity they put out. While they should probably be producing 8 or 9 kw, they are only sending 6 or 7 to the grid. It's very difficult to get an idea of how much this is, and I just realised that the display panel will only record a month of readings, so I only have records from 1st March. The display panel also has a "clear" button, which is usually the button for "select", and it does not ask you whether you are sure you want to clear the data. So it's very easy to lose it all. There is no way to automatically get the data from the display panel to a computer, so I've been copying it by hand.

Anyway, in March, we produced an average of 30 kWh per day, the solar panels were generating an average of 665 minutes (11 hours) and the output was being limited for an average of 20 minutes per day; a total of 6-and-a-half hours over the month. If we guess that it's limiting the output by 2 kW, then that's 13 kWh per month. That's over one percent of the total monthly production of around 1000 kWh. At 48 yen per kWh, that's a large beer.  

Last Saturday (20th April), I put an electric fan in the machine room, pointing at the power conditioners with a timer on the plug so that it comes on between about 11 am and 2 pm. I'm not sure of the effect, or how to measure it, without standing next to the panel all day for the next two years. I'm sure the amount of energy used by the fan will probably average out more or less the same as the amount of extra electricity it will allow to be generated. 


(In case anyone didn't get 2 + 2 = 5,

eg. 2.4 + 2.4 = 4.8

2.4 is a large value of 2, in a scenario where you're rounding to the nearest whole number. Similarly 4.8 is about 5, so 2 + 2 = 5.)

Friday 27 April 2012

The final piece of work - a pricey towel rail

Here is the last piece of contracted work within the house. Apart from a few other bits that we're going to be doing over the next couple of years. And of course the things that were not done properly in the first place and need fixing. 

Technically speaking it's a radiator, but most of the time it's just going to function as a towel rail. We managed to get through the first winter without it even being there, so I'm not sure whether we'll ever switch it on, or if we do, when and for how long. Perhaps it can go on for half an hour on winter mornings to make the changing room toasty after an early morning shower.

One useful feature is that the socket is 200 volts, rather than the standard 100. This means, with a plug adaptor, I can charge my UK mobile phone.

Sunday 15 April 2012

The blind man came to see the house

With our triple-glazed argon-filled low-e windows we don't really need curtains. At least that's what the people selling the windows said. Actually when it's really cold outside, some blinds would help reducing heat loss through radiation, especially with the lack of emmissivity in the room. Also blinds give us a bit more privacy, and may help keep some of the summer solar radiation out, although I think geometry and the balcony will be more effective. Just because we don't really need them, doesn't mean we don't want them.

Anyway, for aesthetic purposes as well as mitigating the flow of heat, we want broad wooden venetian blinds along the ground floor on the south. From a privacy perspective, it's easy to angle such blinds so that it's possible to see out, down to our terrace and garden, without being seen into from the higher windows of the neighbouring buildings.

We also need a blind in the washitsu, the tatami room, for privacy when guests are staying in there. Personally I'm quite happy for anyone who gets to the front door to be able to peer into that room, but this is not a unanimous view among members of the household. I was originally hoping that we'd get a shoji paper door across that window, but at some point it was decided that this would not match the Germanness of the window. Perhaps sliding shoji doesn't go with swinging windows. Perhaps for this reason, perhaps for other reasons best known to the architect, there is very little clearance between the top of the window and the ceiling in this room. Around 40 mm from the ceiling to the top-sill, then 35 mm from the top-sill to the top of that window sash when it opens, inwards. That's only 75 mm, which is not a lot to fit a blind in when it rolls up to the top, as it inevitably will. And I have some memory of the ceiling having been raised, so there would have been much less space to put the blind in.

Perhaps this was in the architect's blind spot. Or perhaps he just turned a blind eye. 

The blind we were looking at for this room was kind of a double roll blind, that has alternating strips of transparent and opaque material doubling back on itself, so, as you let the blind down, it goes from being opaque to having strips you can see through. When I say opaque, it's not completely opaque like air-raid curtains, or the wall next to the window; more strictly speaking it's translucent. The critical part for our concerns is that the width of the roll, when rolled up, is 78 mm, a whole 3 mm wider than the gap between the wall and the open window. Not ideal. Because the blind is doubling back, it is actually twice as long as the height of the window, so a single roll-blind would probably be significantly thinner, and should fit in the space. It wouldn't give us this find level of transparency adjustment though. Another kind of these blinds has two different rolls, with different colours, and presumably each roll would be thinner, but that blind seems only to be attachable in a vertical orientation, with one roll on top of the other. You could say it's a case of the blind leading the blind. 

Onto the South side, the middle big window has a bit of ceiling sticking out on the left side, and a beam coming across on the right side, so we are limited for clearance there too. There is about 150 mm from the protruding ceiling to the window top-sill, then about 40 to the top of the window sash when it opens. The stark choice is either a narrow blind, where there is a few centimetres' gap on each side, or a shorter blind, but the shorter blind would need to be half a metre shorter to make a significant difference to the thickness of the slats when closed. 

One suggestion to get around this was to make the blinds wider than the box with the mechanism in it. It looks like the total height of mechanism plus slats was over 200 mm, but taking away the box width from this, the width of the slats seems to be less than the 150+40 gap we have between the protruding ceiling and the top of the beam. The blind man said he was going to find out from the manufacturers whether they could produce blinds like this, where the slats are wider than the box. He got back later saying that they could make blinds like that, but the whole thing actually needed about 300 mm, so there wouldn't be enough room for the slats. I don't completely believe this, as usual, and would like to see some numbers and actual clearances before a decision is made. I can appreciate their desire to leave wide clearances, and plenty of margin for error, but I also have a strong desire to have blinds that will cover the whole of the window reveal. 

And just in case you thought you'd seen the last of the blind puns, I'm going to blindside you with another one. 

Thursday 5 April 2012

What comes down sometimes doesn't go up again.

From the beginning of February, when it started getting to around minus ten each night, we were putting down the shutters on the upstairs windows each night. They may have some insulation effect, but their cutting down on heat radiating from the house is probably more significant.

Joe, who is eight, correctly surmised that the best time in the morning to put the shutters up is when the light comes on to show that the solar panels are producing electricity, and the shutters should come down in the evening when the light goes out. According to the passive house software, the heat coming in through the south-facing windows is 790 kWh in February, which is over 20 kWh per day. This goes up to 927 kWh in January, over 30 kWh per day. It's 843 in December and 743 in March. Overall, the south facing windows bring in over three times more heat than they lose, so even if it's very cold outside, any solar radiation coming in is going to make up for the thermal losses, and certainly be more of a benefit than any increased insulation from the shutters.

Anyway, I got back one evening to find that one of the shutters had come down a few centimetres. I assumed this had been partially put down, so I switched it to come down properly. It didn't, but there were whirring noises. I switched it to go up. More whirring noises. 

Now this shutter happens to be outside the window opening onto the balcony. Lucky, I thought, as I can just open the window and try to help it up or down. Actually I got Joe to help with the controls, while I stepped out onto the balcony to try pushing the shutter up while he was switching the shutter to close. I thought perhaps it was a little disengaged.

As it turns out, it was more than a little disengaged, and after some more whirring and a little clunking, the whole of the shutter came crashing down to the bottom, leaving me standing on the balcony in bare feet on a very cold afternoon. It was impossible to get the shutter up again, as it was not coiling properly inside its case. The shutter case uses some kind of six-pointed star screwdriver, so I couldn't open that to get the shutter up. I was stuck.

I considered climbing down from the balcony, but decided it was best to appeal to my better half's better nature and get her to bring a ladder out. The sharp tongue of a wife who is both angry and right, while certain to cause some injury, would be much less painful than potential injuries from sliding and landing on the hard tiles below. 

Before fixing this, the suppliers, who also provided our windows, needed to get parts for the disengaged shutters, and there was a chance that they would not be able to fix it until we got back from our travels. They assumed that we would be leaving the shutters down while we were away, but I was looking at the thermal gain from these windows. Of a total almost 16 square metres of glazing on the south side of the house, the shutter in question was covering a double window, with 1.5 square metres of glazing. That's 10%, so a couple of kWh a day. The equivalent of a heater on for an hour. It can still be quite chilly at the end of March, so we didn't want to be turning heat away while we were away.

As it happened, the parts arrived and it was fixed in time. What had happened, apparently, was that the for the motor to stop when the shutter was going up had been set above the actual level of the shutter when fully wound up. The switch goes three ways: up, off and down. If the shutter is down, when the switch is switched to up, it will go up until you switch it to off again, or if you leave it it will go all the way to the top, then it should switch off inside, even if the switch is still set to up. So the first few times we used it, we had obviously been very careful and attentive, marvelling at the wonder of this technology, and staying to switch it off as soon it reached the top. After a week or so, we switched it to go up and left the switch, confident that it would automatically stop. However, it wasn't switching off and the motor was still working away on the shutters, presently rending them from the axle. 

All is now well, and I'm beginning to wonder what temperature it is actually worth putting the shutters down at night, as the effects on radiation must be minimal.

Tuesday 3 April 2012

Luke warm water

It was a beautiful sunny Sunday in February, so we went skiing. There's nothing nicer, after a hard day of skiing, than a nice hot bath. Usually I'd go to an onsen, a natural hot spring, and soak in abundant mineral rich waters, not worrying that they are probably heated with dirty oil because they lose so much heat to the outside air.

There was school the next day, and anyway, we have this fantastic new bath which can supply hot water, and it's only a short crawl to bed from there, rather than a drive. 

I just went with the kids, so I called as we were setting off asking the wife to put the bath on. We didn't get on that you could telephone.

When we got home, looking forward to a piping hot bath to sooth are muscles, the bath was not really all that hot. The first sign of something wrong was that the control panel didn't have the red light on to show that it was keeping the bath water warm. It's set to keep it warm for a couple of hours.

The previous day, we'd used too much hot water in the underfloor heating in the morning, and so there was no hot water in the evening and we didn't run the bath. That meant that it was starting from cold. 

I pressed the button again, and it announced that it was going to start running the bath, and the red light came on, and the little bath icon with lines floating up. A few minutes later it just stopped. No warning, no error message and no bleeping. It just gave up, put the lights out and the icon vanished. There were three bars when we'd got home, and these were starting to go down to two and then to one. 

When I checked the bath, it was not getting much hotter, but did seem to be getting fuller. 

The problem is the strategy for filling a bath. 

If the bath is empty, it's easy. It just needs to fill the bath with how ever many litres we've told it, at whatever temperature it's set for. Simple calculation based on the temperature of water coming out of the tank. 

Usually the bath is not empty though. As we get clean outside the tub, we don't need to change the bath water every day. I guess it measures the amount of water in the tub by the water pressure at the fancy jet thingy where it sends water in and out. If there is less than the required amount of water in the bath, it can get the water to the right temperature by adding hot water. Again this is a fairly simple calculation.  100 litres at 30 degress needs 50 litres at 60 degrees to make 150 litres at 40 degrees.

If the water in the tank is not hot enough to do this, then it can go to plan B, which is to circulate the water from the bath through the boiler, and re-heat the water in the bath. This is known as Oidaki.

The problem with this strategy is if the water in the boiler isn't hot enough, it will just take away heat from the boiler, and still leave the bath luke warm. Remember the second law of thermo dynamics. 

So I suspect what's happening is that the Eco cute tries to add hot water to the bath, realises that it can't get it hot enough with hot water, then tries Oidaki, and then gives up.

We just have to make sure that there is enough hot water, and in this case there wasn't for about four reasons.

First of all, the night before had been very cold, minus 11 in the morning, and not ideal conditions for an atmospheric heat pump. Exactly how unideal is an ongoing concern.

Second, we'd left the underfloor heating on for rather too long in the morning and run off a lot of the heat from the boiler. This left it in a situation of abundant water at low temperature.

Third, we had a full bath of cold water that had not been heated the previous day. Part of the reason why it  was full is that the washing machine is not working properly. It's supposed to be able to reuse the bath water in part of its cycle, but doesn't seem to be working. The bath water, controlled by a special tap and with a separate hose into the washing machine, keeps running into the washing machine when the tap is on, while the washing machine should actually be switching it on or off itself. That's another story though.

Fourth, and this may have been fairly critical, the temperature of the boiler was set to low. I'm not sure why, but I remember showing the wife how the temperature setting worked a couple of days ago, and maybe I didn't set it back again. If the boiler had been set to a higher temperature, in fact, everything would have probably been OK. It is now.
at least influencing 

Sunday 1 April 2012

Difficult decisions... Counting the number of angels that can dance on a pinhead

February 19th. One of those really cold mornings after a bitter snowy day that was followed by a starry night, and weather that seems to have come straight from the Arctic. Minus eleven when I got up and looked at the data logger in my room that's connected to the outside thermometer. It was under 14 on the thermometer on the window sill, strategically placed in the coldest spot in the house.
The panels were already making 0.7 kW just after 7 am, highly efficient supercooled by the ambient temperature and then some by radiating beyond the stratosphere with nothing coming back.
And I wanted to make a cup of tea.
Usually I turn the IH stove onto a high middle setting to boil the kettle. There are ten bars, and I'll put it to number seven. I think gas stoves are most efficient at a middle setting. They may boil the water quicker if you turn it right up, but they will use more energy to do so. I assume the same for IH heaters, although they may be equally efficient at any level.
Anyway, I was thinking about all those lovely kilowatt hours, and wanting to sell as many of them as possible. Putting the kettle on at any level was going to exceed the 0.7 kW we were generating, and mean buying electricity. Given this, the logical thing to do was to turn the IH stove up as high as possible, and while it was on, put the shutters up, which use a couple of hundred Watts, to keep the time that we were buying electricity short, and we were back to selling electricity as quickly as possible.
This must have saved at least some fraction of a yen.
These are the kinds of calculations that we are forced into by the economics of solar power. Surely I have better things to do. The best thing, of course, would have been to drink water rather than tea.
I did notice, after switching the kettle off and setting off proudly with my tea, that we were still using 400 Watts, which seems strange when everyone's still asleep and nothing's on. I realised it was the pump for the underfloor heating, which I'd set to come on from 7 to 7:30 as well as an hour before 6. This was because I'd left it on too long the previous morning and it had used up all the hot water, so it didn't come on in the evening, as I'm mean and didn't let it start working till cheap electricity rates kicked in at 11pm.
With the sun already beating down, we aren't going to need heating until tonight, so it's a good thing I noticed it. That will actually have saved a few yen.