Out-of-Calibration CRM and Wood Stove Spikes
You are hired to test a two-story 1962 home with a basement wood stove, a sump pit, and a CRM whose calibration sticker is dated 14 months ago. The homeowner wants a 48-hour test before listing. You place the CRM in the basement rec room, 20 inches above the floor, and start the test Tuesday. The homeowner uses the wood stove both evenings. Hourly data show 2.0–2.4 pCi/L during the day, spikes to 9.6 and 8.8 pCi/L during stove operation, and a return to about 2 pCi/L afterward. The 48-hour average is 4.6 pCi/L. Your backup CRM is committed to another job until Friday, and listing photos are scheduled for Saturday. The homeowner asks whether the stove should be shut down for the "official" test.
1. What is the problem with the CRM's calibration status, and can you use the data?
Read the worked answer
The CRM's calibration is 14 months old, and annual calibration is the baseline requirement for defensible CRM data. A device past its calibration interval should not be used to produce a reportable result, even if it appears to be functioning; the data are not traceable to a current calibration and can be challenged. The correct action is to stop the run, document the calibration status, and restart with a calibrated device - your backup on Friday, a rented or borrowed unit, or a charcoal or alpha-track device if the timeline allows. Do not back-date or rely on the previous calibration. If you already collected data, keep it in your file but mark it non-reportable and explain why to the client. Scheduling pressure does not make an out-of-calibration instrument acceptable.
2. Interpret the spike pattern. Would the 4.6 pCi/L average be the right number for decision-making even with a calibrated device?
Read the worked answer
The spikes line up with wood stove operation, which is a classic pressure-driven pattern: the stove and flue depressurize the basement, pulling more soil gas through the slab and sump, and the readings fall back when the stove cools. The arithmetic average of 4.6 pCi/L is only meaningful if the test itself is valid, and here the calibration issue means it is not. Even with a calibrated monitor, a 48-hour average can hide short high episodes; the hourly pattern is important because it shows the home has a route for soil gas and a driver that can worsen it. Do not tell the homeowner to stop using the stove for the "official" test - that would test a condition the family does not live in. Instead, test under normal occupant behavior and consider a long-term test to capture the seasonal average.
3. What recommendations do you give the homeowner, and how do you handle the non-reportable run?
Read the worked answer
Recommend a new 48-hour test with a calibrated CRM as soon as one is available, and a long-term alpha-track test (at least 90 days, ideally covering the heating season) to characterize the annual average. Recommend that the sump pit be sealed with a gasketed cover and that the wood stove installation be checked for proper make-up air and draft, because the spike pattern suggests depressurization. If confirmed results are at or above 4 pCi/L, recommend a mitigation evaluation; if they are between 2 and 4 pCi/L, discuss options and retest. Put all of this in writing with the caveat that the original run is non-reportable because of the calibration lapse, and keep the original data in your records.