TO: CHAIRPERSON JOHNSON AND MEMBERS OF THE AD HOC COMMITTEE ON WOODSMOKE

FROM: EXECUTIVE OFFICER

: SUBJECT: FIREPLACE "EQUIVALENCE" IN THE MODEL ORDINANCE

---

Executive Summary

At its August 27, 1998 meeting, the BAAQMD Board of Directors Ad Hoc Committee on Woodsmoke considered a draft model woodsmoke ordinance that could be adopted by Bay Area cities and counties. The ordinance would apply to the installation or replacement of wood-burning devices within a locality adopting the ordinance.

In developing specific language for the model ordinance, the Committee asked staff to (1) determine whether it is possible to establish a particulate matter (PM) certification standard for fireplaces that is "equivalent" to the standard for woodstoves found in EPA's wood heater new source performance standards (NSPS) and (2) if so, provide language for incorporating this "equivalence" into the model ordinance being considered by the committee.

To respond to committee's request, staff conducted an extensive review of the technical literature on fireplace emissions; met with Mr., Paul Tiegs of OMNI Environmental Services, Inc., the company that has done all fireplace testing for compliance with the State of Washington fireplace standard; and discussed fireplace emissions and standards with other air quality agency staff. This memorandum addresses the committee's request and conveys the following staff findings and recommendations:

Findings

• Fireplaces and woodstoves are not equivalent devices. A typical fireplace produces much more particulate matter (PM) per hour or in a typical evening's use than a typical woodstoves because of the fireplace's higher burn rate (amount of wood burned per hour). "Equivalence" between standards for fireplaces and woodstoves is not possible.

• The so-called "equivalence" for fireplaces developed by the Washington State Department of Ecology is an equivalence Mandated by statute, The Washington standard does not result from the identification of existing or potential emission control technology and the setting of an appropriate standard based on those control technologies, as is normally done in setting an air pollution standard.

• The Washington State "equivalent" standard for fireplaces appears to favor small fireplaces over larger fireplaces but does not distinguish between cleaner fireplace technologies and dirtier fireplace technologies.

• There are no proven fireplace designs available at reasonable costs that produce significant reductions in fireplace emissions.

• The Washington State fireplace test protocol appears to be adequate, but the protocol for sampling PM emissions allows sampling to continue well after a fire is out, thus producing misleading emission rate calculations.

• Inadequate data exists to demonstrate the repeatability of test results for a particular fireplace when tested by the same lab or by different labs using the Washington State lab method.

Recommendations

• The model ordinance should not include "equivalence" language for fireplaces.

• The model ordinance should not reference the Washington State fireplace standard or test method.

• The model ordinance should be modified to require that all wood-burning devices be EPA-certified units or pellet heaters, or be gas-burning devices. The requirement that wood-burning devices be EPA-certified would mean that builders who want to install wood-burning devices would have to install EPA-certified zero clearance "woodstove" fireplaces ("EPA fireplaces") or pellet heaters instead of non-certified "builder boxes." Where gas units are installed, provisions should prohibit conversion to non-complying wood-burning devices.

A revised model ordinance incorporating these recommendations is attached.

Fireplace "Equivalence"

The NSPS - Emission Rates vs. Emission Factors

In 1988, EPA adopted New Source Performance Standards (NSPS) for woodstoves (1). EPA projected that the NSPS would reduce woodstove emissions by about 70 percent (1). However, the EPA standards apply to only a small subset of wood-burning devices in common usage. By carefully defining and limiting the applicability of the standard to small woodstoves, EPA was able to set an emission rate limit which prohibits particulate emissions greater than 7.5 grams per hour.

An emission rate limit Like the one in the EPA standard is somewhat unusual in air pollution control. Most emission Limits are concentration limits (e.g. 15 ppm NOX for a gas turbine), emission factor limits (e.g. grams of VOC per gallon of coating applied), or control requirements (e.g. reduce emissions by 90%). These three types of limits do not impose any restriction on the size of an operation (e.g. a turbine, a coating line, or other operation). They simply require that emissions, regardless of overall amount, meet the limits or restrictions.

An emission rate limit, like the one in the EPA NSPS, has a tendency to require different levels of control depending on the size of the source. A small source might produce 7.5 grams per hour of emissions without controls and would therefore require no additional controls. A large source on the other hand, might produce emissions far greater than 7.5 grams per hour, thus requiring significant combustion controls like staged combustion or post-combustion controls like a catalyst.

For the woodstove NSPS, EPA limited applicability to small sources by defining an "affected facility" as a wood heater (woodstove) with (1) an air to fuel ratio of less than 35 to 1, (2) a firebox volume of less than 20 cubic feet, (3) a minimum burn rate of less than 5 kilograms per hour (kg/hr), and (4) a weight of less than 800 kilograms (1). It then set an emission rate standard that could be met by stoves using combustion air controls or catalysts. At the time, EPA recognized that such an emission rate standard had no useful application beyond the narrow universe it had defined, but argued that the standard allowed woodstove-to-woodstove comparisons, which was all it sought (2).

Fireplaces

The NSPS does not apply to fireplaces or other devices that do not meet the definition of "affected facility" in the NSPS (1).

There are four types of wood-burning devices that are commonly called "fireplaces" and that are not subject to the woodstove NSPS:

• Traditional masonry fireplaces. These are the traditional site-built fireplaces constricted of masonry. Though normally built from bricks, several manufacturers now offer factory-made cast masonry core components around which a masonry fireplace can be constructed. Two such devices are the Frisch Rosin and Buckley Rumford fireplaces. Masonry fireplaces are extremely expensive and are generally found only in high-end new construction. In addition they raise significant seismic concerns which add to cost.

• Zero-clearance fireplaces. These are metal fireplaces designed to be installed into wood framing. They are sometimes called "factory-built fireplaces" or "builder boxes." They are open fireplaces and do not meet the EPA definition of woodstove because they have an air to fuel ratio greater than 35 to 1. They are cheap and extremely common in new construction. Nationally, 80% of the 27 million U.S. fireplaces are zero-clearance fireplaces.

• Masonry heaters. Masonry heaters are a traditional northern European means of heating using a small open firebox set in a massive masonry structure with horizontal and downward flues though which gases are channeled before reaching the chimney. Masonry heaters store the heat from intermittent rapid fires and radiate, it back to the building. Unlike the three categories above, masonry heaters may have relatively hi thermal efficiency and may be used for heating purposes rather than aesthetic or recreational purposes.

There are two types of devices commonly called fireplaces that are subject to the woodstove NSPS or, in the case of some pellet-burning devices with air to fuel ratios greater than 35 to 1, that can meet the NSPS emission limits:

• Certified fireplace inserts. These are woodstoves, including some pellet stoves, designed to be retrofitted into a traditional masonry fireplace.

• Certified "EPA fireplaces." These are woodstoves designed to be installed into wood framing in the same manner as a zero-clearance fireplace. Unlike a zero clearance fireplace, these devices meet the woodstove definition in the NSPS.

In addition to the above wood-burning devices, there are two types of gas-burning devices that are called "fireplaces." They are not subject to the NSPS.

• Gas-burring fireplaces. These are factory-built fireplaces that are available as inserts or zero-clearance models and with several different venting arrangements.

• Gas logs. These are gas burners installed in a masonry or factory-built fireplace to replicate the look of a traditional wood-burning fireplace.

Fireplace Emissions

EPA compiles emission factors for sources of air pollution in a document called AP-42. The AP-42 emission factors are used by local and state agencies in calculating emission inventories, The AP-42 particulate emission factor for wood-burning fireplaces is 17.3 grams per kilogram of wood burned (3).

EPA's emission factors are based on limited data. Given the significance of fireplace emissions in many local PM emission inventories, it is surprising how few studies have examined fireplace emissions. The following table (derived from Jaasma (4) with data adding two more recent studies) sets forth results from the available fireplace studies.

STUDY	YEAR	NO. of FIREPLACES	AVERAGE FUELING RATE	AVERAGE PM EMISSION FACTOR (g/kg)	AVERAGE PM EMISSION RATE (g/hr)
Reitz (5)	1993	1	7.9	8.6	79.7
Jaasma (6)	1992	1	3.3*	10*	32**
Jaasma (4)	1992	6	3.7	13	44
Colorado/Shelton (7)	1987	1	3.5	15	53
Dasch (8)	1982	2	4.1	8	33***
ARB (9)	1980	2	15.7	33	518***
EPA/PEDCo (10)	1979	2	5.1	21	107***
EPA (11)	1975	1	7.3	10	76

Notes

* Calculated from other data in report
** Excludes kindling phase emissions
*** Calculated by Jaasma based on other data in report

The more recent studies listed above involved explanation or validation of emission test methods. Because dry fuel, generally dimensional lumber, was used, the emission factors and rates are probably lower than they would be in real-world fireplace use. However, though there is limited data, the more recent studies indicate that fireplace particulate emissions under test conditions appear to be about 50 grams per hour and about 12 grams per kilogram of wood burned.

In the June, 1998 issue of HPA Journal, a quarterly journal published by the Hearth Products Association, Dr. James Houck, vice president of Omni Environmental Services, Inc. states that "... the typical emission rate of a simple open radiant fireplace is about 60 g/hr"." Omni has done all of the compliance testing to date for the Washington State fireplace method.

In a meeting with District staff on October 8, 1998, Paul Tiegs, also of Omni, discussed fireplace emissions. In a paper he prepared to summarize information on fireplaces, Mr. Tiegs stated:

"The emission factor for fireplaces given in AP-42 is 17.3 g/kg, This value was estimated by the EPA using limited field and laboratory data. Estimates from more recent (albeit also limited) data, produces a value of about 12.5 g/kg. The typical burn rate of a fireplace is 3 kg/hr. The emission rates corresponding to an emission factors of 12.5 g/kg and 17.3 g/kg with a burn rate of 3 kg/hr are 37.5g/hr and 51.9 g/hr. There has been some data suggesting that 60 g/hr is a more representative fireplace emission rate. Clearly, the particulate emission rate from a fireplace is dependent on what the typical burn rate is. There has been no consensus of opinions among industry experts on how reasonable these emission values are for fireplaces. Some think they are too low, some think they are too high and some think they are reasonable. There is, however, general agreement on the fact that t he data base is very small and emissions are variable. (13)"

Comparison of Fireplace Emissions to Woodstove Emissions

In addressing fireplaces in the model ordinance, it is important to know how emissions from fireplaces and woodstoves compare. To compare particulate emissions between fireplaces and EPA-certified woodstoves, it is useful to know how the two are typically used in the Bay Area. Very few homes in the Bay Area rely solely on wood as a heating fuel. According to the 1990 census 41696 Bay Area units rely on wood for heat. given an estimated 300,000 woodstoves in the Bay Area, most woodstove use is for recreation and auxiliary heating. Though the District has no survey data on this point, it is reasonable to assume that typical woodstove use is heating a living area for an evening starting at 5:00 pm and ending at midnight.

There are an estimated 1 million to 1.5 million fireplaces in the Bay Area. Although there is also no survey data for Bay Area fireplace use, a Colorado survey and a survey by Dennis Jaasma at Virginia Polytechnic Institute indicate that the typical fireplace fire lasts for 4 hours (6). Based on these assumptions, the table below compares emissions from typical fireplace use and typical woodstove use. Note that emission factors and rates for both woodstoves and fireplaces are based on laboratory testing. Actual emissions in real-world use may be greater for both woodstoves and fireplaces. Emissions from pellet heaters and gas fireplaces are included for comparison.

DEVICE	EMISSION FACTOR (G/KG)	EMISSION RATE (G/HR)	LENGTH OF EVENING USE	PM EMISSIONS IN A TYPICAL EVENING OF USE
Wood-burning fireplace	12	50	4	200
Certified woodstove	7.3	7.5	7	52.5
Certified Pellet heater	2.1	1	7	7
Gas fireplace	n/a	0.07	7	0.5

Note

* Calculated from AP-42 emission factor for natural gas combustion of 3.38 grams per million BTU of gas combusted (15) and assuming heat input of 20,000 BTU per hour.

Fireplace proponents occasionally claim that fireplaces burn "cleaner" than woodstoves. There is an element of truth in this. Though they produce more particulate pollution in typical use, some fireplaces can burn more cleanly than some woodstoves if you look at emission factors rather than emission rates. In addition, fireplaces can also produce significantly lower polycyclic aromatic hydrocarbon (PAH) emissions than even the cleanest certified woodstoves (except for certified pellet heaters). PAHs are a subgroup of polycyclic organic matter (POM). POM is regulated under the federal Clean Air Act as a hazardous air pollutant and by the State of California as a Toxic Air Contaminant. A number of PAHs are carcinogenic. The EPA emission factor for all fireplace POM is 0.008 Fares per kilogram of wood burned (3), For certified catalytic stoves, however., the PAH emission factor is 0.20 grams per kilogram and for certified non catalytic stoves, it is 0.25 grams per kilogram (14). Even allowing for the difference in burn rates, fireplaces can be cleaner than woodstoves for PAH. However, as noted earlier, emission rates for combined particulate matter are much higher for fireplaces. Note that pellet heaters burn more cleanly than fireplaces or woodstoves and have a PAH emission factor of 0.0001 grams per kilogram (14) and gas fireplaces are significantly cleaner than any wood-burning device.

"Equivalence' - G/Kg vs. G/Hr

The use of the term "equivalence" in describing the Washington State fireplace standard or other proposed fireplace standards is unfortunate and misleading. Despite their similarities, fireplaces and woodstoves are not equivalent devices. Woodstoves are lowburn-rate devices with burn rate controlled by restrictions on combustion air. The average of the four bum rates used in EPA test procedures used for wood stoves is 1.03 kg per hour. Woodstoves typically have thermal efficiencies (a measure of the ability of a device to transform fuel into heat) of 60-70% and are used for heating.

Fireplaces are high-burn-rate devices (i.e. they burn a lot more wood in an hour than woodstoves) that have unrestricted combustion air. Fireplaces generally have burn rates of about 3 kg per hour and low thermal efficiency (7-11%). They are generally not used as heaters and are instead used for aesthetic or recreational reasons.

As discussed above, fireplaces produce much more particulate pollution per hour than woodstoves, largely because of their higher bum rates. As a result, fireplaces cannot compete with woodstoves on an emission rate basis in grams per hour.

Origins of the Washington Fireplace Standard

A 1994 law directed the State of Washington's Department of Ecology to establish statewide emission performance standards for wood stoves and fireplaces (16). The law Stated that factory-built fireplaces were to meet the EPA NSPS standards "... or equivalent standard that may be established by the state building code council by rule" and were to be tested using a method developed by the council. The council was also directed to develop design standards for masonry fireplaces that achieved emission reductions similar to those achieved by factory-built fireplaces.

The Washington Building code Council was thus faced with a legislative mandate to develop an "equivalent" standard for devices (fireplaces) than cannot possible meet the EPA standard. The essential problem with this legislatively-mandated "equivalence" is the difference in the burn rates (and emission). Because fireplaces burn much more wood per hour than certified woodstoves, they cannot meet an emissions rate limit for woodstoves. To get around this problem, the council used the AP-42 emission factor for certified woodstoves of 7.3 grams per kilogram. This became the Washington state limit for fireplaces.

Air pollution standards are normally developed in a manner very different from that used to derive the "Washington standard. The first step in setting a standard is to determine the type and magnitude of typical emissions from the source. The second step is then to identify control technologies and the degree of emission reduction achievable by each technology. The final step is to set a standard based on a chosen technology, taking into account the availability, cost, and environmental impacts of the technology (17).

For the woodstove NSPS limit, for example, EPA identified specific control technologies, including combustion controls and catalysts, that are proven means of reducing particulate emissions (1). It then set a standard based on emission reductions achievable using those controls. The Washington state limit, on the other hand, does not result from such an examination. Instead the standard is purely the result of the legislative mandate. It simply translates an emission rate for woodstoves into an emission factor and then applies it to fireplaces.

As a result of the failure to set a standard based on control technologies, the Washington standard appears not to draw a line between clean and dirty fireplace technologies. It appears only that small fireplaces tend to make the list while larger ones do not. since evidence indicates that emissions factors vary with fueling rate (6)=, small fireplaces would tend to have lower emission factors and would therefore tend to pass the Washington emission factor standard. This tendency appears to be borne out in the limited testing done for the Washington standard (18). Thus, it may be that the only effect of the Washington standard is to screen our large fireplaces.

Are There Cleaner Fireplaces?

In setting a fireplace emission standard, the question should be, "Are there any technologies for reducing emissions?" At present the answer is "Possibly, but only the extremely wealthy can afford them." According to Dr. Dennis Jaasma, a national expert on fireplace emissions and testing:

In reality, fireplace field emissions are virtually unknown and the existence of one or more clean-burning fireplace designs in the field is a matter of speculation(20).

Several "fireplaces" on the Washington list use identified control technologies, either combustion air controls of catalysts, to reduce particulate emissions. However, these units are EPA-certified woodstove "fireplaces". Most of the units on the list, however, are "builder boxes" of conventional design with no control technologies. Two cast masonry systems on the list, the Buckley Rumford and Frisch-Rosin systems, for which emission reduction claims have been made, performed more poorly than a conventional masonry fireplace in tests by Dr. Jaasma(20).

In discussions with District staff, Paul Tiegs admitted that he did not know whether there is a cleaner-burning fireplace(18). He said that some extremely wealthy homeowners in Aspen, Colorado, which has strict woodsmoke controls, have installed electrostatic precipitators in masonry fireplaces. Electrostatic precipitators are extremely expensive both to install and to operate. Dr. Jaasma has apparently worked on an afterburner for fireplaces, but, according to Tiegs, there are significant difficulties with such an approach.

Washington Test Method

To certify fireplaces, Washington State adopted a test method developed by its Building Code Council(21). The method is the only adopted regulatory fireplace test method in the nation.

After review of the test method, discussion with Mr. Tiegs, and discussions with staff from other California air districts, staff feel that the Washington method is sound in most respects. However, staff have identified two potential problems with the method.

The first problem is that the sampling protocol allows sampling to continue well after a fire is out. This is because the method requires sampling to begin when the temperature of the flue gas is 25 degrees F. higher than the ambient temperature and end when the temperature of the flue gas returns to this level. Especially with masonry systems, the result may be a multi-hour period at the end of the test when the masonry is still radiating retained heat, no combustion is occurring, and samples are of clean air. Although this extended sampling period should not affect a test result expressed as an emission factor (i.e. in terms of grams of PM per kilogram of wood burned), it has the effect of lowering the calculated burn rate (i.e. kilograms of wood burned per hour), since no more wood is burned during the extended period. If this lower burn rate is then used to convert the emission factor into and emission rate:

emission rate (g/hr) = emission factor (g/kg) X burn rate (kg/hr)

it will result in a lower emission rate than would be the case if the burn rate were calculated for the shorter period when the burning is actually occurring. this produces misleading emission rate results.

The second problem with the test method id that inadequate data exists to demonstrate the repeatability of test results for a particular fireplace when tested by the same lab or by different labs. As noted, only 14 fireplaces have been tested, and all test were conducted by OMNI Laboratories in Oregon. Without repeated testing of fireplaces by OMNI and by other laboratories, nothing can be said about whether different labs would produce the same ranking of fireplaces by emission levels. Such repeatability of results is important for a fireplace test method(20).

The Northern Sonoma County APCD is now working on a test method that may minimize or eliminate the problems associated with the lengthened test period of the Washington State method(22). Northern Sonoma currently proposes to end sampling when flue gas oxygen returns to a level that indicates active burning has ended. Northern Sonoma also currently plans testing to verify test repeatability.

Recommendations

Based on the analysis summarized above, staff make the following recommendations:

1. The model ordinance should not include "equivalence" language for fireplaces.

Discussion: As discussed at length above, fireplaces and woodstoves are not equivalent devices. fireplace emissions are significantly higher than woodstove emissions and only appear similar when viewed on an emission factor basis. There are no reasonably-available emission control technologies available for fireplaces as there are for woodstoves.

2. The model ordinance should not reference the Washington State fireplace standard or test method.

Discussion: The Washington State standard, unlike the NSPS, is not the result of an analysis of emissions and available control technologies, with the setting of a standard that favors clean technologies over dirty technologies. Instead, the Washington standard is the result of a legislative-mandated equivalence. The fundamental problem with the standard is the lack of any reasonably-available control technologies for fireplaces. "The test method appears to include an appropriate test protocol but the test termination procedure produces misleading emission rate results. Repeatability of test results has not been demonstrated.

3. The model ordinance should require that all devices be EPA-certified units or pellet heaters, or be gas-burning devices. where gas units are installed, provisions should prohibit conversion to non-complying wood-burning devices.

Discussion: Most fireplaces now installed in new construction are "builder boxes". The requirement that wood-burning devices be EPA-certified would mean that builders who want to install wood-burning devices or pellet heaters would have to install EPA-certified zero clearance "woodstove" fireplaces ("EPA fireplaces"). The aesthetic differences between builder boxes and EPA fireplaces are minor. The major difference is cost. The higher costs of EPA fireplaces would reduce builder profits. Gas burning fireplaces would be less costly and more environmentally desirable than wood-burning devices. Where builders install gas appliances, the model ordinance should prohibit conversion to wood-burning devices not allowed under the ordinance.

A copy of suggested changes to the model ordinance to incorporate recommendations is attached.

References

[not complete]
Snowden, W.D., D.A. Alguard, G.A. Swan5on, =d W.E. Stolberg. 1975. Source Sampling .Resi4dential Fir,-placesfor Emission Factor Development. Research T@gle Park, NC: U. S. EPA, Pubt. No. EPA 450/3-76-010. 12. Houck, J.E. 1998. "It's Win-Wia: New Hearth Product Sales Can Be Part of the Solution for New Air @ty Regulations." HPA Journal 2(2); 10-18. 13. Houck, J.E. and P. Tiegs, P.F-. 1998. "Fireplace Use and Tectmology Review." Paper presented to Bay Area Air Q=4ty M=agement District staff, October 8, 1998, 14. U.S. Environmen@ Protection Ageucy. 1996. " 1. IO Residentw Wood Stoves." AP-42, @mpilation OfAirPdllutantEminion Factors: (Extc@ Combustion Souices), 5th ed. Research Tnangle P@ N-C.- USEPA. 15. U.S. Eavirownental Prot@on Agency. 1998. "1.4 Na@ @ Combustion." AP-42, Compilation ofair Pollutant Emission Factors. (Exte@ Combustion Sources), 5th ed. Research T@gle Park-, N.C.; USEPA. 16, Re@ed Code of Washington section 70.94.457. 17. See, e.g., California Health and Safety Code s@ou 40406 de@g "best available retrofit control technology." 18. Tiegs, P.E. Presentation to Bay Area Air @lity Management District staff, October 8, 1998. 19. W@ou State Department of Ecology, "Certified Woodstoves, Authorized Fireplaces and Other Solid Fuel Buming Devices." List maintained by DOE Air Q=hty Pro@ i-@ed October 2, 1998. Olympia: Washington DOE. 20. Stern, C.H., D. R- J and M.R- @ion. 1993. "Masonry Fireplace Emissions Test Method; Repeatability and Sensitivity to Fuelwg Protocol." Environ. Mon. Assess. 25. 51-64. 21. "Washington State Standard Test Method for Particulate Emissions from Fireplaces." Washington Uniform Building Code Standard 31-2, Washington Adminitrative Code 51-30-31200 section 31.200. 22. Northern Sonoma County Air Pollution Control District. 1997. Testing Protocolfor ParticWate Emissions)'rom Masonry Fireplaces. Draft test method dated December 10, 1997. H@dsbuyg, California: NSCAPCD.

Back to Fireplace Emissions

[The Rumford Store]
[Rumford Picture Gallery] [Specifications] [Plans & Instructions]
[Manufacturers] [Dealers] [Architects] [Builders] [Masons] [Sweeps]
[Associations] [Masonry Chimneys] [Masonry Heaters] [Training] [links] [search]
[Superior Clay Corporation]

Buckley Rumford Fireplace Home Page
Copyright 1996 - 1999 Jim Buckley
All rights reserved.