Dripping Diary

26th January, 2016

Water started dripping from the ceiling in the pantry this morning. 

When you have water dripping out of somewhere, it's a good idea to find where it is coming from and stop it from going in there. 

The immediate suspect, like the last five times water has appeared in unwanted paces, was the ventilation system two floors above.

This is actually the second water incident in the last months, but the first one was quickly noticed from the sound of drips on the bathroom ceiling, so it never got to build up anywhere. 

There was no dripping on the bathroom roof this time. That's because it was not the ventilation system leaking, even though that's where the water was ultimately coming from.

I quickly came to a second hypothesis. Half a metre of snow outside... temperatures below freezing for a few days... the drain from the ventilation system coming out of the wall about twenty centimetres above the ground... frozen pipe! 

left: drain from ventilation system

The first evidence to support this hypothesis was the water spilling gently over the the top of the drain beneath the ventilation unit, rather than actually going down it. 

A bit of hosepipe with a loop usually goes from the ventilation system to the drain. I diverted it into a bowl to stop sending more water to the overflowing drain. 

The next evidence was outside: a large icicle coming out of the drain.

A few buckets of hot water and kettles later the icicle was gone. The visible part of the icicle went fairly quickly and was soon followed by the rod of ice from within the pipe. It took a little longer to thaw the elbow. Immediate problem solved, it was time to address the cause.

The ventilation system is going to produce condensate when it's cold outside, unless we also make it cold inside, or drop the relative humidity below about 20%. It's often going to be below freezing when it's cold outside, and that's where the water is going to be dripping. So it seems inevitable that ice is going to form and, sooner or later, the outlet pipe will freeze. It will then fill up with water and start overflowing. The only mysteries are: why has this not happened before? and why did the contractors not prevent this from happening?

I suspect this probably has happened before, but it takes a while for the pipe to fill up with water before it overflows, then it takes a while for the water to drip down, around the bath that is one level below the ventilation system, then onto the ceiling of the pantry below the bath. Some of this water will be evaporating all the time, and it could be a couple of days before enough builds up to break through the plaster boards and start dripping onto the floor. By this time, the temperature outside has always gone high enough above freezing, or a few rays of pre-noon sunshine have reached the drain and thawed it. 

I was wondering if there were any mitigating circumstances leading to this, and there are a couple of things that may have made a difference. I noticed when I was clearing away another icicle a couple of days later that I'd left a gardening stake directly underneath the drain, from which was growing a nice icy stalagmite. Perhaps such a stalagmite had helped to block the drain. I was in too much of a hurry to melt the ice before and didn't document the hydro-crystalline pathology very well. 

The other thing that we had done the night before this incident was to put on the humidifiers. It gets a bit dry in the winter since we're constantly getting rid of our humid air, and replacing it with air that's already fairly dry, and is then being heated so that the relatively humidity will fall about four times. We don't have any permanent remedy for this, but when we remember, and when it gets below about 30 percent, we switch on some of our humidifiers, usually at night time. So we are adding moisture to the warm air that we are expelling from the house over a steep temperature drop, and increasing the amount of water that will end up in condensate.  

According to my previous calculation, the amount of water that's going to drip on a cold night is up to around 450 ml per hour. One drop is 0.05 ml, so that would be about two and half drops per second. Not fast enough to represent constant flow, but perhaps slightly faster than the ideal drip rate for forming icicles, which seems to be around one or two grammes per minute according to the Icicle Atlas. The precise temperature and humidity of the air in the house will determine the dew point, which will likely be a few degrees above zero. The dew point is really the critical number since it will tell us when condensation starts. 

It's 23 degrees C and 30% humidity right now, so the dew point is 4 degrees. (According to this dew point calculator.) So if the air outside goes below about 2 degrees, it's going to drop below the dew point within the ventilation system. If we had 100% efficient heat exchange, then it would be cooling the air all the way down to 2 degrees, and the air coming in would be heated all the way up to 23 degrees. It's more like 80% so we lose a couple of degrees. Some ventilation systems will recover only 60% of the heat, so they will be less likely to reach the dew point. This is only a problem that will happen in well-ventilated houses with highly efficient heat exchange ventilation systems, so I suppose that answers the question of why the contractors hadn't thought about this happening, and why we've had so many problems with this. 

When I say we've had many problems, we haven't exactly been wading through water, just needed to use a small cloth to mop up a few drops from the floor every year or two. And hopefully the structure of the house has not been damaged by the moisture. 

The other problem ventilation systems have to deal with is freezing condensate. If the air is being cooled below freezing, it may start snowing in there as vapour in the air is precipitated. This would block the ventilation and we would no longer be able to ventilate the house, so the ventilation system does something with pressure differences to stop that. I'm not really sure what it does, but the result will probably be that it never gets as low as zero in there, and in fact there may only be a very narrow window of outside temperatures when condensate is actually being produced.

(Apologies to anyone who was hoping for a story about cooking fat from the North of England.)

Note:

The other drain in the picture is from the air conditioner, which we have hardly every used. This has a de-humidifer on it, and it would take moisture out of the hot air if we were using it. There is no chance of it freezing though. 

Read about previous escapades here.

Lesson 13: How do air conditioners work?

I made a bit of a miscalculation. There were three lessons left and three student presentations, one of which was on low-energy buildings in hot climates, and another on generating your own power. I had planned to do a lesson on cooling, and had some leftover material on solar electricity, and should really have had these presentations on three different days, then filled in the rest of the class with my information, assuming that the students had not covered it in their presentations. Or if they had, then continue the lesson with discussions on the relevant topics. Unfortunately they both ended up in week 14, so week 13 became empty, and the only material ready for it was the left-overs on cooling and solar electricity. 

If I'd been doing a full lesson on cooling, I would have started with a brainstorm of different ways to stay cool, with my non-exclusive list ready in the wings: windows, insulation, thermal mass, trees, heat-exchange ventilation, fans, air conditioners, de-humidifiers, and ice—preferably large blocks. Another thing that was not on my explicit list, which probably should be, was shading. 

Not wanting to take away too many options for the following week's presentation, and keen to teach some science, I skipped this and went straight into an exposition of the workings of air conditioners. 

Before getting into the details, we needed to understand four concepts relating to gases: volume, pressure, temperature and heat. These are all interrelated. Other things being equal, if a fixed amount of gas is in a smaller volume, it will have a higher pressure. Other things being equal, if the amount of heat in a fixed amount of gas goes up, the temperature will rise. If you don't add any heat to some gas, but squash it in to a smaller volume, then the amount of heat won't change, but the temperature will go up. And if you expand it, the temperature goes down. 

A heat pump sends a fluid in a circuit through a hot area and then a cold area. The Fluid is compressed as it goes into the hot area, which will increase the temperature and allow it to transfer heat to the hotter area. It is then allowed to expand when it goes into the colder area so the temperature will drop and heat well flow from the cold area into the fluid. 

That's the Carnot cycle. Heat pumps are basically trying to defy the second law of thermodynamics, by getting heat from a colder place to a hotter place. We use them in our fridges and air conditioners, and increasingly they are used for space heating and water heating. 

The coefficient of performance is used to measure the efficiency of a heat pump, and it measures the amount of heat that is transferred divided by the amount of energy that goes in. Typical domestic heat pumps have average COPs of 3 to 5, but precise numbers are very difficult to find. 

There are limits to the COP, and as the temperature difference goes up, the COP will go down. If a heat pump is used for generating hot water in the winter, using cheap night-time electricity, the COP can get very low, and the heat pump is not performing much better than an electrical heating element.  

Lesson 12, part IV: Life cycle analysis

So you have to look at the whole lifecycle, like we did with power generation. We need to look at positive and negative impacts during manufacture, installation, use, decommissioning and disposal.
Let's look at some glass wool insulation. You can get a roll 11 metres long, 910 mm wide and 100 mm thick for around 6,000 yen. For the same price you could get around 90 litres of paraffin. So which should you get?

You need to keep warm, and you can either wrap your house in the glass wool, or burn the paraffin. How long would it take for the insulation to save the energy in the fossil fuel?

The glass wool has a U value around 0.44 W/m2K, and we can assume 80 thousand kelvin hours per year heating demand. There's about ten square metres of it, so putting it on a wall will stop something like 352 kWh of heat per year.

A litre of paraffin has around 9.8 kWh of energy, so the 90 litres have 880 kWh.

Therefore, it will take about two and half years of heating bills to pay back the insulation costs.

We're forgetting a few things in our calculations. Of course we need a heater to burn the paraffin, and we need to install the insulation. It's not going to make us any warmer just by buying it and putting it in the corner.

We're probably not starting from zero insulation, but adding that insulation on top of existing walls. If we already have a U value of 0.44, this extra layer will only save half the heat. The more insulation we have, the less heat there is for extra insulation to stop.

We also need to remember the Jevons paradox. If a house is poorly insulated, and paraffin needs to be bought and burnt every time we want to heat it, it is probably not going to be at that ideal inside temperature all the time.

If we have a traditional building with very little or no insulation, the temperature will be low a lot of the time, and the heating bills will be moderate.

We could put more heating in to get the building to a comfortable temperature, but the heating bills will be very high.

Adding a little insulation will mean the temperature is higher, but we are probably going to be using the same amount of heat we started with to keep it comfortable for longer. This is the Jevons Paradox.

If we get enough insulation, then the temperature can be kept comfortable the whole time, with much lower heating bills.

A little insulation is a dangerous thing!

Back to the insulation and heating oil, we can also look at the carbon costs. While the financial costs were similar, in terms of CO2 emissions, burning the paraffin will emit two hundred times more carbon than manufacturing the glass wool. This puts the carbon return on investment around five days.

In all these calculations we need to look at the trade-off between running costs and lifetime, since the total cost of the house includes initial costs plus running costs multiplied by the lifetime. The length of the lifetime makes a difference to these calculations. If you have a building component with a pay back of twenty-five years, that makes sense in a house that will last a hundred years, but not in one that will last fifteen years. But who would build a house that only lasts fifteen years?

So we have a vicious cycle here where houses are worth nothing after less than twenty years. This means banks give small loans, buildings are built cheaply, they are not maintained and are often knocked down within twenty years.

This short lifetime means that return on investment calculations will prevent investment in long-term energy saving technology.

More serious, short lifetime of building means massively more energy is spent on the buildings. Low energy investments are usually tiny percentages of the total building cost. A building that uses half the energy over its lifetime does not use twice as much energy to make. It may use ten percent more. Often low energy buildings are high technology and while the costs may increase, in carbon accountancy there is no difference.

So if buildings have a lifetime of twenty years, they could be using four or five times more energy than in another country where their life expectancy is a hundred years. This is obvious. The reasons why Japan has a disposable building culture are a little more complicated. How Japan can get out of this situation is the trillion yen question!

References
Ito, Akiko (2013). Policy and programs for energy efficient houses and buildings
Further reading

Life cycle cost and carbon footprint of energy efficient refurbishments to 20th century UK school buildings — ScienceDirect

Further listening
Freakonomics: Why are Japanese homes disposable?

Lesson 12, part III: Economics, the dark side of Energy efficiency

"Like the 2006 changes, it was predicted that the introduction of these limits would result in a 20% reduction in energy use for heating. A survey by Liverpool John Moores University predicted that the actual figure would be 6%"

UK energy policy, laws and regulations handbook: strategic information and basic laws.

Why did a predicted 20% reduction in energy use turn out to be 6%? The 1974 Warren Alquist act in California was predicted to lead to 80% saving in energy efficiency. Research into actual energy use has suggested they made no difference.

Remember that the oil shock inspired low-energy buildings in Europe and energy efficient products in Japan?

Today, country A uses four times more heating than country B, while country B uses fifty percent more to twice as much energy for hot water, lighting and appliances. It seems like country A would be Japan and country B somewhere in Europe. In fact country A is Germany and country B is Japan. In spite of low energy building standards, Germany uses four times more energy for heating, and in spite of all those energy efficient products, Japan uses up to twice as much energy for hot water, lighting and appliances. Or perhaps it is because of the efficiency.

Jevons paradox, inspired by 19th century economist, suggests that more efficiency will lead to more consumption. He was looking at coal, for example Watt's steam engine that was 75% more efficient than its predecessor, but ended up using much more coal.

A good way to think about this is with vending machines. There are plenty around Japan, serving hot and cold drinks to thirsty customers with a couple of coins rattling in their pockets. Considering the economics, we have the cost of the drinks, the maintenance and the heating and cooling on one side, and the income from the money going in on the other. The cost of the drinks will be balanced per drink sold, but the heating and cooling costs are going to depend on the time and the temperature outside, so there is some kind of break-even point of the number of drinks that must be sold per month.

What would happen if the vending machines became twice as efficient? They wouldn't need to sell as many drinks per month to break even. This means there are a lot more places where vending machines could be installed, and the result would be more vending machines. More efficiency would not lead to less energy use, but to more energy use.

These low-energy building standards and highly efficient products may not be helping the energy problem at all. They may be leading to more energy use. Or at best, we may just be using the energy we saved elsewhere.

Engineering estimates don't take into account consumer behaviour.

They can try, as we can see in the next part.

Reference
Freakonomics: How efficient is energy efficiency?

Lesson 12, part II: The future

Here are some quotes about the future, taken from the past.

• "Heavier-than-air flying machines are impossible."
• "This 'telephone' has too many shortcomings to be seriously considered as a means of communication."
• "Who the hell wants to hear actors talk?"
• "Everything that can be invented has been invented."
• "There is no reason anyone would want a computer in their home."
• "I think there is a world market for maybe five computers."
• "640K ought to be enough for anybody."
• "…data processing is a fad that won't last out the year."

The answers are at the bottom, although you'll have to work out the order yourself.

The point is that predicting the future is very difficult, even for very clever people.

In 1943 there didn't seem to be a very big market for computers. A few years later the transistor was invented, and in the 1960s somebody worked out how to put more than one onto the same chip of silicon. Moore's law started to take effect then, leading to a doubling of the number of transistors on an integrated circuit every couple of years, which goes on to this day. These kinds of exponential growth often have finite limits. We hope that fossil fuel consumption and carbon emission will not continue on its exponential growth, and if the amount of installed solar power continues growing exponentially it will reach the total solar irradiation on the entire surface of the earth. It has been suggested that Moore's law will stop working when the size of transistors hits the size of molecules and atoms, but recent research into quantum computing may allow the trend to continue.

Practically speaking, Moore's law tells us that if we wait another couple of years, the cost of an electronic device will halve, or the capacity will double. This could mean that you should wait a couple of years, or it may mean that you should just go ahead and get what you can now, since you would be waiting another couple of years for ever.

LEDs are one example of a future technology that has come into the present in the past few years. Back in the seventies, a "zero heating building" could have reasonably used 100 watt lightbulbs in each room, taking lighting as a given, and the excess heat of incandescent bulbs as free heat. Haitz's law is the version of Moore's law that applies to LEDs, and it means that LEDs are now cheaper than other forms of light as they will use less energy and last longer. They are also small, and don't radiate heat or attract insects so they are good in the summer.

Conventional disadvantages are the inability to dim, limited colours and high cost, all of which are fading away.

The future is bright, but only partly predictable. There is also a dark side.

Quotes by:
Western Union internal memo, 1876.
Lord Kelvin, president, Royal Society, 1895.
Charles H. Duell, Commissioner, U.S. Office of Patents, 1899.
H.M. Warner, Warner Brothers, 1927.
Tom Watson, IBM chairman, 1943
Editor in charge of business books, Prentice Hall, 1957
Ken Olson, Digital Equipment Corp. president, chairman and founder, 1977
Bill Gates, 1981