This article was originally published in the March/April 1993 issue of Home Energy Magazine. Some formatting inconsistencies may be evident in older archive content.



| Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy |



Home Energy Magazine Online March/April 1993




`Read Me Your Thermostat':
Short-Term Evaluation Tools


by Laurence F. Kinney

Laurence F. Kinney is president of Synertech Systems Corp., an energy research and services organization in Syracuse, N.Y.

Installing and using run-time meters the right way, in concert with well-planned weekly calls to meter reading clients, can provide effective evaluation of weatherization work.

It's not easy to measure savings from energy conservation work, yet it is crucial for a job well done. Many of the problems facing evaluations based on fuel savings result from the time required to collect adequate consumption data. Evaluation designers generally believe it's important to know household fuel usage for as much of the heating season (and non-heating season, to establish baseline consumption) as possible prior to weatherization and for as long as possible after that. They consider 18 months an adequate time span for gathering data. In fact, data is seldom collected over an 18-month period. To save time, the crucial before-weatherization data that provides the base from which savings are calculated is usually gathered from records kept for other purposes, such as meter-based utility bills.

An approach which shows particular promise makes extensive use of simple and inexpensive instrumentation to measure quickly--over weeks, not months--how weatherization changes the efficiency of dwellings. This approach lends itself to large numbers of dwellings and is quite suitable for agency staff and others who may not have technical backgrounds. This approach does not measure savings at all, but rather enables one to project what savings can reasonably be expected barring major changes in occupancy or behavior. Utilizing short-term consumption data can, in turn, help estimate appropriate levels of weatherization investment. The tools must fit the approach (see accompanying article, Keeping A Running Score on Weatherization,).

Elapsed Timers

Most furnaces and boilers used for residential heating are either off or on, and when on, they consume energy at a constant rate. Measuring the rate of consumption is fairly straightforward. In the case of natural gas, watching, with a stopwatch in hand, the dials turn on the gas meter while the furnace (and only the furnace) is firing yields sufficiently accurate data on firing rate. For oil furnaces, that rate can be determined with sufficient accuracy by pulling the burner and examining the flow rate etched on the nozzle in gallons per hour. (The Btu in fuel oil or natural gas can vary by up to 3%, so precision is quite difficult to achieve.)

If one knows the rate of consumption of a furnace or boiler, then by measuring the time it is operated over a given period in which outside temperature is tracked, it is possible to produce a simple but useful weather-adjusted index of consumption, Btu per heating degree-day (Btu/HDD). This is calculated by the elapsed time multiplied by the firing rate of the furnace (or boiler), all divided by the heating degree-days for the appropriate reference temperature over the same period. Dividing by the heated area gives the energy intensity (Btu/ft2-HDD).

Over the last 12 years, Synertech has used elapsed timers to determine savings due to both building shell and HVAC system retrofits. Under most circumstances, we prefer simple electro-mechanical devices that use 24VAC (alternating current voltage) synchronous motors to drive mechanical wheels calibrated in hours and tenths of hours. These are widely available for $10-$40 (see box, Short-Term Metering Equipment Makers, p. 37).

Gas Furnaces

In the case of natural gas or propane-fired furnaces or boilers, the elapsed timer is wired in parallel with the gas solenoid valve (see Figure 1). When the current flows in the gas solenoid, the valve opens, gas flows, and the timer begins. When the thermostat is satisfied, the gas solenoid valve closes and the timer quits registering time. Since the timer we recommend is a mechanical device rather than digital, power outages will not cause a loss of data.

Oil Furnaces

In the case of oil-fired furnaces, it is sometimes possible to work at 24VAC , but we prefer to measure when the motor that pumps oil (and air) is actually running. We use a transformer, housed in a small box with a fuse, as an intermediary between the oil burner and the timer. The 120 VAC primary (input) of the transformer is wired in parallel with the oil burner motor, and the 24VAC secondary (output) is wired to the timer. In practice, the transformer is located 6-10 ft from the burner--frequently mounted to a solid joist above the furnace--and is connected to the burner with a flexible, armored BX cable. The timer, on the other hand, is located where the client can read it conveniently--usually in the living area, close to a telephone. We use ordinary thermostat wire to connect the timer to the transformer box (or the furnace in the case of gas). Office staff then can call the client weekly to check readings and calculate the week's consumption for that house using a simple worksheet.

The Thermostat As Data-Logger

A new generation of set-back thermostats have several advantages over elapsed timers. Chief among the advantages is the ability to store the cumulative time that the thermostat calls for heat. Several models produced by the Hunter Fan Co. feature a review function that provides the total elapsed time the thermostat calls for heat today, yesterday, this week, and last week, where a week ends at midnight Sunday. A filter review command totals the time the system has run, up to 1,000 hours of firing time--a feature intended to prompt the homeowner to change the furnace filter, but useful for our purposes as well.

Installing a thermostat is easier than installing an elapsed timer, the readout is more accurate, and the recorded time periods are identical for all users (midnight Sunday to midnight the following Sunday). For weatherization programs, it may be counted as a material expense, thereby reducing the material expense for evaluation to zero. (For disadvantages of using the Hunter thermostat this way, and alternative wiring methods, see box The Hunter Thermostat As Logger--Special Cases.)

Filling Out The Audit Forms

Before The Calling Begins--and In Between

Several things need to be in place before taking the weekly client readings of the elapsed timers or thermostats. Let's look at a typical house assuming that the crew installed Hunter's Auto Temp or Energy Monitor II Plus thermostat (or equivalent) during the audit and that the client agreed to receive weekly phone calls to report the previous week's consumption. (In the case of clients without phones, using a pre-addressed and pre-stamped postcard is adequate.)

On that initial visit, the auditor will begin filling out the form (see Figure 2):

The firing rate (Btu/hr) for furnaces using natural gas is determined by measuring the time for a full rotation of the 1 ft3 or 2 ft3 dial on the gas meter while only the furnace is firing. Assuming that 1 ft3 of natural gas has an energy content of 1,000 Btu, the auditor determines the cubic feet per hour of gas flow which then is multiplied by 1,000 to yield the firing rate in Btu per hour. The auditor compares the firing rate, determined by the timing method, to the label on the furnace. If there is a significant difference, the procedure is repeated.

The fuel dispersion nozzle in an oil-fired heating system is inscribed with the rate at which the nozzle will allow oil to flow in gallons per hour (for a pump pressure of 100 psi). The flow rate multiplied by the Btu capacity of the fuel (approximately 140,000 in the case of No. 2 grade of fuel oil) gives the Btu/hr firing rate of the system.

Firing rates for furnaces using propane are read directly from the label on the furnace itself.

The house floor area (ft2), used to calculate energy intensities, should include only deliberately conditioned areas, not the basement or attic unless each is heated.

Heating degree days (HDD), should be determined and recorded on a calendar each day, using a reference temperature of 65deg.F. A local weather station can provide daily readings, but other sources such as newspapers are also adequate. The problems with National Weather Service data are the four-month delay in delivery of the data, the possibility of a significant distance between the audit site and weather station, and the unwieldy format of the data on the diskette provided by the weather service. (Synertech Systems Corp. offers a special weather station with accompanying software that updates local heating degree-day files automatically.)

The point is to have a consistent, reliable source of temperature data from a location close to the monitored dwellings. The average temperature for a 24-hour day (the mean of the high and low is satisfactory for these purposes) is subtracted from 65 (if the reference temperature is 65deg.F) to determine the heating degree-days in a specific day. Only positive values are used; in other words, if the average temperature is higher than 65deg.F, there are no heating degree-days that day.

The auditor should fill in the sum of heating degree-days from last Monday through the most recent Sunday (column 10) before telephoning the client, and this number should be stored in the memory of a hand calculator.

Over The Phone--and Afterwards

The auditor obtains other information during the phone call to the client (columns 1-7), with the remaining calculated after the call (columns 8, 9, 11, and 12).

Although it is not essential for calculating weekly consumption, we recommend that the client be asked to read out the current room temperature, and to then push the filter review button briefly to read the total firing time since reset or installation. Temperature information is useful in tracking client thermostat-controlling behavior. The caller can use total information from the filter review command to reconstruct a missed week of information. One drawback of the Hunter thermostat and logger is that it displays information for 10 seconds after one pushes a button, but holding a button down for more than 3 seconds erases the data!

The information in the last column, Btu per HDD per ft2, should be monitored with each weekly call to spot irregularities. Variations of more than 30% from week to week should prompt double checking of calculations, and then calling the client to verify the reading or check on any changes in lifestyle. (Did the furnace quit working? Were the grandchildren around the house for most of the week?) If circumstances appear out of the ordinary in the household, it may be wise to exclude that week's data from the analysis, but only if a clear explanation for the anomaly is recorded. Using this type of audit form with weekly calls thus has the effect of performing important quality control checks on the data base, as well as preemptively solving individual mysteries.

What Works

Our experience has shown that elapsed timers coupled with human contact are the most important ingredients of the short-time approach. The basic equipment for this approach includes:

  • A run-time meter which can be installed in virtually any kind of heating system--the type with mechanical movements are preferable to avoid data loss.

  • Alternately, a Hunter thermostat.

  • BX cable, thermostat wire, and 24V/120V transformer.

  • Stopwatch to time metered fuel flow and calculate firing rate.

  • A volt-ohm meter, to identify the right wires and to avoid accidents.

  • A telephone and a hand-held calculator, for when client data-reporting begins.


The caller should be well prepared before the first telephone call to the client, with the audit form primed with information for weather-normalization, firing rates, and heated area of the house. With information on run-time reported by the client, the caller can calculate the weekly energy intensity, before and after weatherization takes place, and can make short work of evaluating the cost-effectiveness of the retrofit work.


Figure 1. Schematic of elapsed-time meter installation to measure furnace firing time. Meter indicates 8 hours, 43 minutes of furnace on-time.


The Hunter Thermostat As Logger--
Special Cases

The Hunter set-back thermostat can used as a short-term monitoring tool, but it may not be appropriate in some circumstances. In some cases, alternatives are possible, in others not:

Some people are intimidated by digital thermostats. In this case, another monitoring technique necessary, like installing an elapsed timer with a remote read-out in the living area.

A furnace hits its high temperature limit and turns off prior to the thermostat being satisfied. Since the thermostat counts the time that it calls for heat, in this case it will overestimate energy consumed by the furnace. The auditor should fix this problem on the initial visit by repairing a defective fan motor, broken belt, or dirty heat exchanger, or increasing the air flow across the heat exchanger by cleaning a dirty filter or unblocking ducts.

The heating system operates on millivolts. Millivolt systems are usually found in old gas-fired furnaces or boilers. The pilot light heats a thermopile that produces about 0.7VDC (direct current voltage), used in a series circuit with the thermostat and solenoid valve to activate the heating system. Thermostats that control millivolt systems cannot use anticipators, and thus do not operate the heating system as efficiently as thermostats that include anticipators. (The Hunter has an electronic anticipator set for a 3deg.F dead band.)

We recommend retrofitting millivolt furnace systems to operate on 24VAC (alternating current voltage) as an energy saving measure. It requires a 24V transformer wired in a series circuit with the new thermostat and the coil of a relay. The poles of the relay substitute for the switch on the old thermostat. Thus, when the thermostat calls for heat, it closes a relay which in turn closes the low-voltage system circuit, opening the gas solenoid valve. Honeywell makes a millivolt system conversion package--a transformer/relay combination that mounts on a standard electrical box--for around $20, and installation is fairly simple.

Electric heat is controlled by a single thermostat. Electric furnaces work just like oil- or gas-fired furnaces. Be sure that the firing rate is constant before bothering to install a Hunter thermostat. (The firing rate is usually listed in kilowatts on the furnace label; multiply by 3,413 Btu/kWh, a standard conversion factor, to get Btu/hr.)

Caution: Whenever working on thermostats associated with electrical heating, use a volt-ohm meter to verify the presence of 24V, not 120V which can shock or electrocute the unaware installer!

Some thermostats associated with cheap but common electric-fired systems are designed to operate at line voltage, 120V, but the Hunter thermostats are designed for only 24V operation. Convert the old thermostat to the Hunter by using a transformer and relay, similar to the conversion for millivolt systems.


The dwelling has electric heat with multiple thermostats. Instead of installing a new thermostat for estimating consumption, rely on utility bills.

A hot water boiler serves a single zone. In a system like this, the thermostat doesn't fire the boiler directly, but rather activates a circulation pump that moves water from the boiler to the radiators and back. The boiler fires when its water cools to the point where an aquastat calls for heat. Thus, the run time of the circulation pump is proportional to the firing time of the boiler, but it is not identical to it. One would expect the differences to vary with outside air temperature and with the length of firing cycle, but the variations may not be large. If they are small, then it should be possible to determine weatherization savings by measuring the water circulation pump on-time using the energy usage feature on the thermostat.

In order to study the possibility of measuring savings in systems of this kind, it's easy to install some simple instrumentation to measure the differences. This would consist of a Hunter thermostat and a combination elapsed time meter and event counter hooked directly to the solenoid that fires the burner on the boiler. It also requires a homeowner who is willing to read three dials instead of just one.

A hot-water boiler serves more than one zone for space heating. We recommend not bothering installing a new thermostat.

The house has a condensing furnace. The operation of a condensing furnace (all new furnaces of 78% AFUE and higher, except pulse, use induced fans and have a pre-fire purge) can cause errors in measuring consumption with the thermostat method. Typically, when the thermostat calls for heat, instead of firing immediately, a small fan switches on to pull fresh air through the heat exchanger prior to opening the gas valve. This purge cycle lasts up to 45 seconds, so the firing time recorded by the thermostat overstates reality by as much as 45 seconds, multiplied by the number of firing cycles between readings. But since typical firing time is about 10 minutes, the error is less than 10%.

For purposes of calculating percentage savings due to weatherization, this problem may be safely ignored, since savings are given by the ratio of consumption before and after conservation. However, since after-weatherization furnace run-times should be slightly shorter, the effect of ignoring the purge time in the calculation should understate (to a minor degree) estimates of percentage savings. Absolute savings figures would also be in error. Therefore the new thermostat should be installed on all furnaces with induced draft fans. (It is important to install event counters on a sample of the monitored homes so that a good working number for the number of firing cycles versus firing times before and after weatherization may be derived from a representative sample of dwellings.)


Interaction Between Timer and Anticipator

There can be a subtle interaction with the thermostat when an elapsed timer is wired in parallel with the solenoid. The timer raises the overall current that flows through the thermostat when the furnace calls for heat. The result is the anticipator producing slightly more heat than it produces without the timer in the circuit, thereby decreasing the firing time of the furnace.

What is the anticipator? Here's what we wrote in Home Energy:

The anticipator in most conventional thermostats is in fact a little resistance heater located very close to the temperature-sensitive bimetallic strip itself. It warms the space around the thermostat element.... The anticipator fools the thermostat element into believing that the room temperature is warmer than it really is, so the furnace is switched off.

The anticipator is necessary because of thermal mass, [in other words,] hot air in the ducts (or hot water pipes), heat remaining in the heat exchanger, and an uneven distribution of heat in the house...The anticipator turns off the furnace before the room air temperature reaches the desired setting because, with the heat from addition thermal mass evenly distributed, the room temperature will stay within the limits of the desired temperature. If the thermostat simply waited until the room temperature reaches the thermostat settings before shutting off the furnace, the residual heat in the system would result in coasting to well above the desired temperature. (From Anticipators: Completing the Feedback Loop, HE Mar/Apr '87, p. 30)

ln practice the additional anticipator heat is insignificant, particularly if one picks an elapsed timer which draws only a small amount of current. However, it's worth measuring to be certain, and adjusting the anticipator if needed.


Short-Term Metering Equipment Makers



Elapsed Timers
Kessler Ellis Products
120 First Ave.
Atlantic Highlands,
NJ 07716
Tel: (800)631-2165
Fax: (908)291-8097

Newark Electronics
4801 N. Ravenswood Ave.
Chicago IL 60640
Tel: (312)784-5100
Fax: (312)907-5211

Hunter Fan Co.
2500 Frisco Ave.
Memphis, TN 38114
Tel: (901)745-9222
Fax: (901)745-9385

North Fork Retrofit
511 Carpenter St.
Greenport, NY 11944
Tel & Fax: (516)477-2922

Redington Counters Inc.
130 Addison Rd.
P.O. Box 608
Windsor, CT 06095
Tel: (203)688-6205
Fax: (203)688-1591


Millivolt Furnace Retrofit Kit
Honeywell Inc.
1985 Douglas Dr.
Golden Valley, MN 55422
Tel: (800)328-5111


Figure 2. Weekly run-time audit form



Related Articles

Keeping a Running Score on Weatherization (Hill)
Measuring the Performance of the National Energy Audit (Sharp)
Moisture and Mobile Home Weatherization (Tsongas)
Profiles of Multifamily Weatherization Projects: A Tale of Five Cities (Kinney, Wilson, and MacDonald)
Selecting an Infrared Imaging System (Snell)
Ten Highly Effective Weatherization Programs (Brown and Berry)
Weatherization Assistance: The Fuel Oil Study (Ternes and Levins)
Weatherization Assistance: The Single-Family Study (Brown and Berry)
The Best Boiler and Water Heating Retrofits (Lobenstein and Hewett)
Combustion Safety Checks: How Not to Kill Your Clients (deKieffer)
Condensing Furnaces: Lessons from a Utility (Beers)
Don't Force Air, Go with the Flow (Springer)
Downsizing Steam Systems (Gifford)
Fireplaces: Studies in Contrasts (Hayden)
Integrated Heating and Ventilation: Double Duty for Ducts (Jackson)
The Key to Persistence (Nolden)
Remodeling Bathrooms: Let the Energy Savings Flow (Johnston)
Making Low-income housing Affordable: The Northgate Retrofits (Patullo)

| Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy |


Home Energy can be reached at:
Home Energy magazine -- Please read our Copyright Notice


  • 1
  • NEXT
  • LAST
Email Newsletter

Home Energy E-Newsletter

Sign up for our free monthly

Harness the power of

Get the Home Energy