Solar Hot Water Performance Measurement Part 2

   With almost two years worth of data from the solar hot water system, now might be a good time to review how well it is working.

   The solar hot water system is run in series with a traditional oil furnace so that the better the solar system works, the less oil is consumed.  For nearly three years now, the run time of the furnace has been measured, a time span that includes one year before the installation of the solar hot water system and two years since its installation.
  
    The furnace is monitored using a Hobo U9-004 motor on/off data logger.  The logger comes with a magnetic back that allows it to be attached to the furnace fuel pump motor.  A sensor inside the logger detects the varying magnetic fields that leak from the AC motor when it is running and creates a log entry every time the motor turns on or off.  This data has to be processed with an Excel spreadsheet to compute "duty cycle", i.e. what fraction of the time the motor is on during the day.

    One last trick for this measurement is the knowing that the burner nozzle dispenses oil at a fixed rate.  Conveniently, my burner nozzle outputs 1 gallon per hour.  So if my furnace runs for 1 hour per day or a 4% duty cycle (1/24), then 1 gallon of fuel will be consumed per day.  At about $4/gallon, that would work out to about $4/day.  So on to the data.

 The graph above shows a full year's worth of data (365 data points) from the 2010 to 2011 heating season.  There are a few things worthy of note. First is that during the heating season, there is a lot of day to day variation in the percentage of time that the furnace has to run in response to day to day changes in weather.  In contrast, the fuel consumption in the summer is more consistent as it only varies with the amount of hot water used and idle losses in the furnace.

   The discussion of the oil consumption is facilitated by constructing a simplified model of the furnace usage and fitting curves based on that model.  The first assumption is that there are two seasons for a furnace, Winter and Summer.  During the winter, the furnace usage depends on 1) home heating load, 2) hot water usage, and 3) idle losses.  During the summer, the furnace usage only depends on the last two of these items.  The next assumption is that the winter usage can be represented by a simple parabolic curve fit and the summer by a constant value.  These lines are shown on the graph.

 The graph above shows three years worth of data, and the impact of the solar hot water heater can easily be seen by comparing the summers.  The summers of 2012 and 2013 show far lower furnace duty cycle  than in 2011.   Ideally, the consumption would be zero in the summer, but unfortunately, the way the furnace controls are set up, the furnace works to keep itself at 140°F all year long whether it needs to be on or not.  This might be fixable in the future with a "furnace bypass" system.

   Also note in the graph that there is significant variation during the winter from one year to the next.  This is most likely due to the intensity of the winter weather varying significantly from one year to the next.   The impact of the solar hot water system in the winter is likely to be small (due to less sun) and difficult to measure due to the fact that most of furnace run time is going into heating the home.

   The graph above shows the same data with a different vertical scale (gallons of oil consumed per day).  In the summer of 2011, about 1.13 gallons/day is consumed.  However, in the summer of 2012, only about 0.47 gallons per day are consumed due to the addition of the solar hot water system.  Over the roughly 150 days of the summer, this resulted in 177 gallons consumed in 2011 versus 71 gallons in 2012.

    This amounts to approximately 100 gallons of oil saved in the summer alone.  At $4/gallon, this means there is about a $400/year savings just based on the summer data.  Now given the over $8000 out-of-pocket expenses I have laid out for the solar hot water system, it might take a long time to break-even on my investment.  But there are a couple of bright spots.  Firstly, there is likely some additional savings in the winter.  Secondly, I should be able to put a furnace-bypass system in place now to further reduce oil consumption in the summer.  Thirdly, it seems likely that oil prices will rise in the future and that would reduce payback time still more.

   The relatively small financial saving here are sobering.  However, solar power systems are too cool to be judged solely on the dollars.  There is great personal satisfaction from heating your home water using the only the sunlight that falls on your property versus from oil being pumped up out of the ground in the middle east and shipped halfway around the world and then burned in your furnace creating local pollution.

Measuring Performance of the Solar Hot Water System

    "So how is your solar hot water heater working out for you?" I am often asked.  This simple question often has me struggling for a simple answer.   For my solar electric panels, I can tell people that the panels make more electricity in the summer than I use and I get a ZERO dollar electric bill about six months of the year.  This seems to satisfy most people.  So why can't I say something similar about my solar hot water panels?  In one word, it's instrumentation.
   Solar electric panels typical record the number of KiloWatt Hours (KWH) produced either daily, weekly, or cumulatively (most likely all of the above).  Solar hot water panels, in contrast, record absolutely nothing.  The differential temperature controller turns the pump on and off as needed, but does not record anything.
     It is technically possible to record the amount of energy produced by the solar hot water panels.  Caleffi makes the WMZ-G1 Energy Heat Meter (257202A $250).  But to use it, you must also buy the Vortex Flow Sensor (NA15015 $117) and the Relative Pressure Sensor (NA15014 $90), both of which also measure temperature.  And if you want the data stored, you must also get the DL2 data logger (257201A $410).   All of which ends up being about $1000 after getting all the fittings to plumb it into your system.  As an Engineer, I love data, but adding $1000 to the price of my system just to get data, seems a little steep.

   Having forgone the expensive option during installation, I put a simple data logger on my 2012 Christmas wish list which my loving wife picked up for me (nothing says you care more than a data logger for Christmas).  This beauty was an Onset Computing Hobo U12-006 four channel 12-bit recorder ($113) capable of recording 43,000 data points all while being powered with a single CR2032 watch battery which lasts more than one year.  Complementing the data logger was a TMC6-HD temperature sensor ($39) which was installed in the bottom of the 80 gallon hot water tank.

 

Unfortunately, the data logger cannot record the amount of energy produced, but it can give some very useful insight into the performance of the solar hot water system.  At this point in time, I have a full year and a half worth of data recorded that can give some graphical feedback on the performance of the system.

A Day In The Life

   The graph below shows a typical 24 hour cycle for the solar hot water system.  Starting at midnight, the bottom of the tank is at 92.7°F showing that there is very little heat left in the tank, although the top of the tank is probably considerably hotter and that is what really counts.

   By around 7 am, the temperature has slowly dropped to 91.0°F due to a small loss of heat from the well insulated stainless steel tank.  Then someone wakes up to take a shower.

    By 7:30 am the temperature in the tank has dropped to 86.0°F as a result of the shower.  Most likely, the hot water from the tank was not hot enough for a shower, and some supplemental heating via the oil furnace was needed.  Nevertheless, the water was at least partially heated by the sun.

    Then the sun gets down to business.  Between 9 am and 3 pm, the 80 gallon tank is heated to 158.8°F.  This is pretty impressive with water more than hot enough for a shower.  Over the course of the afternoon, the temperature of the tank drops to 147°F likely due to some incidental use of water (hand washing, dishwasher, clothes washer, etc) and heat traveling up to the top of the water tank leaving the colder water at the bottom.

   From about 7pm to 9pm, the water temperature drifts slowly downward from 147 to 145.3°F likely due to losses through the insulated tank walls or conduction out through the connected copper pipes.

    At about 9:30pm a second family member takes a shower dropping the temperature at the bottom of the tank from 145.3 to 128.7°F.

    Starting at midnight, the cycle will repeat itself the next day in the same way, as long as the sun is shining.

A Year and a Half In The Life

 The graph below shows about 70,000 data points for the temperature at the bottom of the tank.  Remember from the graph above, that each day the tank cycles from cold to hot to cold again as the water is heated up and then used in showers.  So within a one day cycle it is possible for the measurement to be as low as 60°F and as high as 180°F (which is the maximum setting before the heat is sent to the heat dump).

    Concentrate of the lower part of the graph for a moment.  The horizontal axis starts at January 2012, passes through a full year to January 2013 and on for an additional half year until July 2013.  It is clear that in the winter, the incoming water temperature is about 50°F.  In the spring and the fall, the incoming water temperature increases to about 60°F.  In the middle of the summer of 2013, the temperature seldom falls below 80°F.  However, it would seem unlike that the incoming water temperature is so high.  Instead, the more likely explanation is that not all of the 80 gallons of hot water in the tank is needed in the summer and some residual heat is left in the tank on most summer days.

      Now concentrate of the upper part of the graph.  In January, the water is only heated up to about 75°F by the sun.  If you are a glass is half empty sort of person, it would seem that the solar hot water system is "doing nothing".  But in fact, the water is being slightly "pre-heated" before the fossil fuel furnace has to heat it the rest of the way to 140°F.  So some solar heating is better than none.  

     Starting some time in March or April, more serious water heating begins with temperatures pushing to the 140 to 150°F range.  Also notice that sometimes the sun does not shine for several days and the water temperature remains at 60°F.

    In the middle of the summer, the peak temperatures for the day are often above 170°F supplying ample hot water for showers.

    While this data is interesting, it really doesn't supply a quick answer to the question of "how is your solar hot water system working?"  But at least it shows that the systems is working, and, in the summers at least, working quite well.