Development and Evaluation of a new Depressurization Spillage Test for Residential Gas-Fired Combustion Appliances - Final Report

Summary & Introduction

A new combustion depressurization spillage test for residential combustion appliances has been developed. The test has been designed to accurately measure the amount of combustion spillage from residential combustion appliances and their venting systems when they operate at selected levels of depressurization. The new test uses carbon dioxide (CO2) that is produced in the fuel combustion process as a tracer gas. The test method has been designed so that it will not require the use of precision space conditioning facilities during the test.

During this project seven (7) gas-fired appliances were purchased and installed to evaluate both the new combustion spillage test and the appliances. The appliances were chosen to cover a cross-section of the types of gas-fired equipment that are commonly installed in Canadian homes.
The sample set comprised:

Two power-vented storage-tank water heaters
One code-compliant “mid-efficiency” furnace (with the approved supplemental side wall venting kit that would be required for most if not all “Category 1” negative-vent pressure appliances)
Two high efficiency “condensing” furnaces
Two “direct-vent” gas fireplaces

The products and their venting systems were purchased from regular HVAC distributors and they were shipped directly to the testing laboratory by the distributors. Each product was installed in a test room that had been equipped with an exhaust fan that allowed the room to be depressurized to selected levels. Each of the products was installed, following the manufacturer’s certified installation instructions and using the maximum equivalent length and type of venting materials specified by the manufacturer. The condensing furnaces are approved for installation as either direct-vent or as “single-pipe” systems. They were installed as “single-pipe” units, drawing their combustion air from inside the depressurized test room.
For each unit, tests were initially performed with the test room depressurized by 50 Pa (0.2 inches H2O) compared with the pressure outside the room. If the combustion spillage exceeded 2%, the test was repeated with the room depressurized by 20 Pa (0.08 inches H2O). Finally, if the measured spillage exceeded 2% at 20 Pa, a test was performed with the room depressurized by 5 Pa (0.02 inches H2O).
Each appliance was operated for a five minute period of burner operation with the room depressurization level controlled at the selected value. The burner fuel consumption, the concentration of CO2 in the test room, and the exhaust fan flow rate were monitored throughout the five minute combustion period. Measurements were continued for two minutes immediately following burner shutoff to ensure that any transient spillage of combustion products into the test room that occurred after the burner shut off would be included in the test. Similarly, the exhaust fan was operated to produce the required depressurization level prior to activating the appliance to ensure that any ignition-related transient spillage that occurred at the start of the test cycle would also be included.
For each test, the amount of CO2 that was released into the test room from the appliance and its combustion venting system during the test cycle was determined from the measurements. This was compared with the amount of CO2 that would be produced by combustion of the fuel that was consumed during the test. The ratio of the two provides a direct measure of the combustion spillage of the appliance and its venting system during each test.
The 50 Pa depressurization tests results were quite interesting:

Three products had essentially undetectable levels of combustion spillage
Three products had low, but measurable combustion spillage (between 0.7% and 1.5%)
One product had significant combustion spillage (approximately 13%)

Looked at from a different perspective, if one assumes a performance benchmark for a combustion appliance to have less than 2% combustion spillage at its rated depressurization level, six of the seven appliances that were evaluated with the new spillage test would have “passed” at 50 Pa. The seventh product would not “pass” at either 50 Pa (13% spillage) or at 20 Pa (4% spillage). It would “pass” the test at 5 Pa, at which pressure there was no detectable combustion spillage.
The detailed test results are published in a separate laboratory test report ¹. The results are summarized in Table 1.
This report is intended to provide some of the background for this project and to discuss the test concept and assumptions. It also includes a detailed description of the test procedure that incorporates what was learned during the laboratory testing project. The updated test procedure now includes SI units in the calculations.
¹ Laboratory Evaluation to Assess a proposed Test Method to Determine Transient Combustion Spillage, Bodycote Materials Testing Report 04-06-M278b for NRCan, July 2005