Building control systems rely on the information provided to them by the various sensors throughout the building. Sensors for temperature, light level, carbon dioxide (CO2), and enthalpy (or total energy content of air) are just a few examples. If the critical sensors in a building are inaccurate (that is, significantly out of calibration), the building will not work efficiently, costs will increase and comfort issues will result. Let’s look at how to calibrate your building control sensors, what to calibrate, and a few examples of why control sensor calibration is so important.
Proper Calibration Technique
To calibrate a sensor correctly, you need to use a test instrument that itself has been calibrated according to the standards of the National Institute of Standards and Technology (NIST). This ensures that the instrument is, in fact, accurate within a certain margin of error based on a standardized, repeatable test procedure. There is no use in calibrating one device to another if the baseline instrument hasn’t been properly calibrated itself.
What to Calibrate
You don’t need to calibrate all the sensors on a project by any means. Sensors that are locally adjustable by a user, like a digital thermostat, can be excluded if necessary. The user is going to find a number that’s comfortable so it doesn’t really matter if the T-stat says 72 degrees but it’s really 70 as long as the user is content. You do need to calibrate the critical sensors – those whose inputs are responsible for energy management decisions by the building’s control systems.
Examples of Critical Building Control Sensors
Take, for example, a CO2 sensor that controls ventilation to a conference room. I’ve personally calibrated CO2 sensors from reputable manufacturers that have been over 200 parts per million (PPM) off. I haven’t just seen this problem once or twice — I’ve come across miscalibrated sensors dozens of times in my experience. Building codes require minimum ventilation rates be provided to occupied spaces and measuring CO2 is a way of indirectly quantifying the amount of ventilation being delivered. Let’s assume a building code requires 20 cubic feet per minute per person (CFM/p) of ventilation. The corresponding space CO2 set point would be 930 PPM, assuming ambient CO2 of 400 PPM and office-type occupancy. If that CO2 sensor reads 200 PPM low, the resulting ventilation rate will be roughly 15 CFM/p, or 25% deficient. If it reads 200 PPM high, the resulting ventilation rate of approximately 32 CFM/p is 60% high with a commensurate increase in energy expenditure to condition that air.
Certainly, many sensors come out of the box within calibration, but each sensor has to be set up properly by the controls technician. Typically, you’ve got four choices – 4 to 20 mA (milliamps), 0 to 5 V (volts), 0 to 10 V, or 2 to 10 V. The technician has to make the right choice within his or her software. On a recent project I measured the return air relative humidity (RH) for an air handler to be 54%. The controls system was reading 27%. You guessed it – the sensor was set up in the software as a 0-10 V sensor when it was, in fact, a 0-5 V sensor. What’s the big deal? That sensor is used to calculate the return air enthalpy. It’s then compared to the outdoor air enthalpy and the controls system makes a decision on whether to recondition the return air (called recirculation) or get rid of it and use 100% outdoor air (called economizer mode). The one with the lower enthalpy wins. Had this set up error not been caught that unit would have spent many more hours in recirculation mode rather than economizer mode, with much lower realized energy savings than possible.
I’ve worked with commissioning providers and engineers alike who are proponents of skipping this critical step in verifying a building’s performance because of the perceived savings in fees associated with the time it will take to do the calibrations. In both the examples above, the time to calibrate each sensor is between 5 and 10 minutes. Typical calibration of all the critical sensors on a job site takes approximately one to two hours. Compare the annual savings potential with your or your consultants’ hourly rate and you can quickly see how cost-effective it is to calibrate your building’s critical sensors.