Furnance Duty Cycle

   The primary motivation for putting in the solar panels has been to reduce oil consumption and the ultimate success or failure will be judged by the oil savings.

    But measuring oil consumption is slightly difficult and very low resolution.  The truck comes to your house and fills the tank about 8 times per year. It is difficult to know how much oil is used in any given month and you can forget about seeing any consumption data on a daily basis in a way that is easily possible for metered products like electricity.

   However, I have found another way to measure oil consumption.  The technique has two parts.  First, the furnace supposedly uses 1.0 gallon per minute when it is on.  Secondly, there are data recorders (data loggers) that can measure when a motor is on.

   The data logger that I used is called a HOBO from manufactured by a company called Onset.  This is a great product.  You can log one year's worth of data.  It only uses one CR2032 watch battery, attaches to the motor with a magnet, and remotely senses the motor state based on fluctuating magnetic fields.  It has a USB port and the data can be downloaded to the HOBO software.  It is about $100 for the data logger and another $100 for the software. 

  What was a slight pain was parsing the data in Excel.  What I really want to know is Duty Cycle, or the percentage of time that the furnace is on in a given day.  What I start with is a bunch of time events that indicate when the motor turned on or off.  I spent a few hours writing a VBA macro in Excel to parse the data the way I wanted.  The biggest problem is splitting events that happen just before or just after midnight.  But the result is just what I wanted.

The chart is a little complicated so let me walk you through it.  On the horizontal axis is time, about one year starting at the beginning of heating season around October 7.  The black dots are data points for the duty cycle.  If the duty cycle was 50% it means that the furnace was firing half of the time.  I have almost a full year's worth of data.
   Since the data is a little noisy, I fit it with two curves.  One curve is parabolic and represent the heating season.  The second curve is linear and represents the summer season.  This gives a clearer representation of the two uses of the furnace 1) domestic heat, and 2) domestic hot water.  It seems reasonable to assume that the domestic hot water is fairly consistent throughout the year.  In fact, it looks like there is a baseline usage of about 5% or roughly 1 gallon of oil every day.
    The oil usage might rightly be broken down into three usage
1) Domestic home heating
2) Domestic hot water
3) Idle losses
Even if no one was in the house and it was summer, my furnace would still consume oil due to the old fashion "tankless" water heater.  The tankless heater requires the furnace to maintain temperature 24/7/365 just in case someone wants some hot water.

    It is difficult to separate hot water usage from idle losses so I have grouped them together in one group in the graph.  Hot water and idle losses represent about 40% of the oil consumed or 377gallons.  Of that, probably half is hot water and half is idling losses.  So a big win for the solar hot water system will be for it to make enough hot water to allow me to shut off the furnace in the summer.  If oil is $4/gallon, then 377 gallons/year means $1508/year, but I will never achieve that much savings because it is not sunny every day and I am not sure how to automatically shut the furnace off and I might also need to bypass the furnace to get this system to work, something I have yet to figure out how to do. Step-by-Step.

  For domestic heating, an estimated 555 gallons (60%) are used per year.  So perhaps I should have spent my money on a more efficient furnace rather than solar water heater system.  But then again, perhaps I can do the furnace upgrade in the future.

  The estimate shows about 932 gallons/year total oil consumption.  This number is quite good because in the 2009-2010 season I used 942 gallons, so the estimate might be within +/-5%.  The insight gained from the data logger information more than makes up for the possible +/-5% error.

   Hopefully in a year, we will have lower duty cycle numbers to share with you.