Field Measurement of Multifamily Boiler Losses

Space heating offers the best potential for retrofits in this sector for several reasons. First, space heating accounts for approximately 72% of end use gas consumption in 5+ buildings. Second, space heating is provided by central heating systems in over 90% of all the 5+ buildings in the Minneapolis (ibid) and in almost all of these cases, the heating costs are paid by the building owner. Third, many improvements to the building shell are either physically impossible or well beyond owners' payback criteria. Experience nationwide shows median paybacks for heating system measures ranging from 1 to 9 years. By comparison, median paybacks For shell measures range from 11 to 23 years, in those buildings where they are feasible (BECA-MF database, Goldman et al, 1988).

No systematic research has been done to determine which types of boilers have the greatest losses, how the losses are divided among major loss paths, or how they can be reduced. The seasonal efficiency of multifamily boilers quite literally has been measured in only a handful of buildings: Lawrence Berkeley Laboratory's 1988 review of existing research (Modera) cited measurements for only five boilers nationwide. The limited data that are available show seasonal efficiencies covering a very broad range from 79% (Robinson et al, 1988) down to 50% (Modera et al, 1986). Without better information, it is impossible to estimate the economics of boiler replacements or retrofits with any degree of accuracy, either for designing programs targeted at common classes of boilers or for conducting energy audits in specific multifamily buildings.

This report describes a variety of boiler loss measurements conducted on four multifamily boilers at three different sites. One of the sites has a steel fire tube boiler with a power draft burner and venting that is connected directly to a masonry chimney. This site also has a front end boiler that has copper water tubes with an induced draft burner and venting that is also directly connected to the same chimney. The second site has a steel fire tube hydronic boiler with a power burner and has a barometric damper on the vent. The third site has an atmospheric, site built steam boiler that has barometric and electronic vent dampers.

The test procedures and analysis methods are first presented for three indirect and five direct measurement methods. The indirect methods consist of two techniques used on an isolated boiler (i.e. no loads are applied): the stop loss and heat up and cool down. The third, time to make steam, is applied to a steam boiler while it is operating to meet the building loads. The direct methods for measuring the burner input and stack efficiency are relatively easy to perform. Directly measuring the off-cycle flue loss rate, jacket loss rate, and boiler output is typically more difficult and expensive. The results from the field tests are presented and, when available, data from two different methods on the same boiler are compared. Some of the practical and theoretical limitations of the methods are also discussed.

The stop loss measurements on two of the boilers showed a strong relationship between the stop loss factor and outside temperature. For both of these boilers, the stop loss factor doubled as the outside temperature dropped from 70°F to -19°F. The results From the input/output and time to make steam methods also showed some deviation from those expected from the constant off-cycle loss, linear boiler model. The heating season efficiency was estimated for each boiler. The two hydronic steel fire tube boilers had efficiencies of 80.7% and 77.7%. The copper water tube boiler had an efficiency of 86.1% and the steam boiler 71.6%.

Full Report (PDF)
Field Measurement of Multifamily Boiler Losses