**EDIT, March 2011 **
The modelling that we did with WATSUN (and upon which this blog post was based) was completely inaccurate. It hasn't been near 100%. In January our tank sits at 20 degrees, and it fluctuates between 25 and 40 during February. Not even close to what the software promised us.
I'm not sure what happened, but modelling with WATSUN was a complete waste of time for me.
Our solar fraction is probably in the range of 60%-70%.
The Mill Creek NetZero Home (MCNZH) will collect solar energy in three different ways: using passive solar design, using photovoltaic (PV) modules, and using solar hot water (SHW) collectors. The SHW collectors are the ones that heat water - they are the black ones at the top of the above picture.
According to a recent article in Home Power magazine (Oct/Nov 2008, p.40), SHW collector efficiency is 50%-70%. That's pretty good when you consider that the best PV module is about 17% efficient.
I've been contemplating the design of the MCNZH's SHW system for months now. The pieces started to fall into place once the federal government released a crucial tool: The WATSUN 2008 SHW System Simulator.
Computer Modeling - WATSUN 2008
The ability to model real-life systems with computer software has widespread application in every area of technology. In residential home building, though, I suspect that it's given short thrift. In my experience, rules of thumb and "what worked last time" are what steer the boat when designing residential homes. I prefer to trust in reliable, time-tested computer code more than gut feelings.
WATSUN 2008 is available from the Canadian Federal Government as a free download right here. It performs hourly simulations of SHW systems, and it's based on algorithms that have been in use since the 1970s. One of its great features is that it contains weather and sunlight data for Edmonton, Alberta. Once I downloaded and learned about the software, I contacted Didier Thevenard. He is the primary mover and shaker behind WATSUN 2008, and he generously helped me figure out some of the collector data numbers that I needed.
WATSUN 2008 - Results
I had some pretty important questions to answer when I began the computer modeling:
- How much of my hot water can be produced from solar?
- Should we install 2 collectors or 3?
- How big should our storage tank be?
- How important is pipe insulation (the pipes that run from the collectors to the storage tanks)?
- Is there any energy left over for space heating?
So I set up some simulations. The collectors will be of the flat plate variety, purchased from Trimline Design Centre. They manufacture the collectors right here in Edmonton.
In the first set of simulations, I was testing the effect of three different variables: pipe insulation (R3 vs R6), tank size (316 litres vs 1000 litres) and number of collectors (2 vs. 3). Here are the results:
The bars represent the energy, in kWh, that it will take to heat our domestic hot water.
This first modeling run really impressed on me how important the pipe insulation is. Two Collectors, with a 316 litre tank and R6-insulated pipe (the red dotted line) is virtually as performant as 3 Collectors with a 316 litre tank and R3-insulated pipe (solid blue line). That is amazing, and another testament to the absolute necessity to spend on insulation first.
My simulations got more refined. For one thing, I settled on some hot water usage assumptions:
- There will be 2 adults and 2 children/teens living on the main and second floors, and one adult living in the basement suite (if zoning allows), for a total of 3 adults and 2 children/teens in the home.
Other Assumptions (thanks to Gordon Howell for some of these numbers):
Usage assumptions are very tricky to decide on, but I think these are reasonable and conservative. So the hot water load has increased from the first modeling run to about 149 litres of hot water per day. Here are the results of the second run:
This simulation focused on tank size and number of collectors. I should point out that the 316 litre tank is insulated to R28, and the 1000 litre one is insulated to R50. That's because we would be purchasing the 316 litre one off the shelf with an R8 rating. I am assuming that we would only be able to add a (fairly arbitrary) R28 insulation value. The 1000 litre tank would be home-made, so we would have full control to insulate to R50.
Too Good To Believe?
The best system, with 3 Collectors, a 1000 litre, R50 hot water tank, and R10 pipes, is projected to generate 94% of my annual hot water from the sun. That is an amazing amount of hot water production for a system that is done right. While it does seem quite difficult to believe that more than 90% of my hot water could be produced by the sun in February, for example, I am going to trust my modeling on this one. After all, I believe that the average system is very poorly done, with R2 pipe insulation and R8 tank insulation being the norm. If that's the case, then maybe the rule of thumb that we in Edmonton can only get 50%-60% of our hot water from solar is incorrect. Even if WATSUN is off by 10%-20%, the results would still be impressive.
The final modeling run is trying to answer the question: Is there any energy left during the cold months to heat the home with?. The one place in the MCNZH that may have a problem staying warm using passive solar energy is the basement. In the basement, the windows are partially blocked, so the tenant may be using a lot of electric baseboard energy to stay warm. Could the solar hot water system dump some heat into the basement?
I told WATSUN 2008 that the hot water load would include 62 litres of 40-degree water per hour, for four hours, every afternoon. This extra hot water load represents kWh of heat that we would use (in winter) if they were available.
A lot is going on in this graph (sorry it's so small). The green bars are the energy load without the extra heating (domestic hot water only), and the lighter bars are with extra heating load (248 litres per day, at 40 degrees Celsius).
The last three lines in the legend demonstrate that the 1000 litre tank is perfectly sized to three collectors. You can see in the graph that 2000 and 3000 litre tanks make almost no difference in the final amount of heat delivered. In fact, when we need the heat most in January, the bigger tanks are less performant than the smaller tanks. Also, 316 and 500 litre tanks (at R50) aren't quite big enough, although they are competitive with the 1000 litre tank.
The most important thing that this graph tells me is that there is indeed useful space heat to be harvested from a SHW system in Edmonton. The MCNZH as a whole only needs heat from November to March, but the basement may need extra heat even in October and April due to its lack of windows and the tendency for air to rise. The numbers behind this graph say that the SHW system can provide the following amounts of energy:
- January: 110 kWh
- February: 155 kWh
- March: 238 kWh
- April: 220 kWh
- (summer doesn't matter)
- November: 125 kWh
- December 78 kWh
- Total: 927 kWh
Obviously, the results need to be taken with a grain of salt. Even if the SHW system produces half of this heat, though, it might be worth spending, say, $1000 to include a way to dump heat into the basement when there's a surplus.
My WATSUN modeling results have enabled me to make the following decisions about my SHW system:
- insulate the pipes to at least R6, preferably R10
- insulate the storage tank to R50
- install a 1000 litre storage tank
- 3 collectors
- there is extra heat - install a system to harvest it
Note: I've attached the excel spreadsheet with the last two graphs in it for easier reading. It's here.
(cross posted at raisingspaces.com)