Research civil engineer, U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, NH
Pages: 19 Published: Jan 1987
As-built drawings and handbook calculations of R values are often inadequate bases for investment decisions regarding improved insulation of U.S. Army buildings. Reported field and laboratory experience indicates that a technique employing surface-mounted heat flux sensors (HFSs) in conjunction with infrared thermography (IRT) can yield reliable estimates of R values. This technique employs IRT to position HFSs and thermocouples at representative locations on walls and roofs or attics to acquire heat flow and temperature data for estimating R values. This paper reports on the application of this technique at Ft. Carson, Colorado, and Ft. Richardson, Alaska, to eight family housing units, a temporary office building, and a barracks.
Infrared thermography of these buildings detected few thermal anomalies, but measurement of several walls with HFSs and thermocouples (typically at six locations spaced vertically on each wall) revealed significant variation in estimated R values; this variation is attributable to convection, even within fully insulated walls. This is significant for proper placement of sensors and indicates that installed fibrous insulation can lack the ability to quell convection. The insulating ability of walls containing poorly installed mineral fiber batt insulation was much worse than would be indicated by the design handbook values. Some attic insulation performed exactly as expected; some was at least 40% worse than expected.
Plotted curves using the life-cycle cost assumptions required by the U.S. Army's Energy Conservation Investment Program show the budget necessary for each level of insulation improvement to break even over a 15-year economic life. At Ft. Carson, filling empty stud cavities in brick veneer construction to an overall R value of 13 would warrant an investment of about $17.6 per square metre ($1.63 per square foot). A contractor estimates that the actual cost would be only $4.5 per square metre ($0.42 per square foot). At Ft. Richardson an R-20 level of insulation would justify an expenditure of between $12.9 and $30.1 per square metre ($1.20 and $2.80 per square foot) of wall on the wood frame buildings that were measured. Attic spaces at both installations were heavily insulated and had much less potential for improvement than walls.
The IRT/HFS measurement technique is most useful on poorly insulated buildings, where the inaccuracy of information about the initial insulating level has the greatest impact on the allowable budget for additional insulation.
heat flux sensors, in situ R, value measurement, heat flux sensor calibration, measurement error analysis, building thermal diagnosis, natural convection, radiation heat transfer, conductive heat transfer, insulation retrofit, energy conservation, economic analysis, thermal insulation
Paper ID: STP18483S