# 10 CFR § 429.18 - Residential furnaces.

§ 429.18 Residential furnaces.

(a) Sampling plan for selection of units for testing.

(1) The requirements of § 429.11 are applicable to residential furnaces; and

(2) (i) For each basic model of furnaces, other than basic models of those sectional cast-iron boilers (which may be aggregated into groups having identical intermediate sections and combustion chambers) a sample of sufficient size shall be randomly selected and tested to ensure that -

(A) Any represented value of estimated annual operating cost, energy consumption or other measure of energy consumption of a basic model for which consumers would favor lower values shall be greater than or equal to the higher of:

(1) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(2) The upper 97 1/2 percent confidence limit (UCL) of the true mean divided by 1.05, where:

$\begin{array}{c}UCL=\stackrel{—}{x}+{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the number of}\phantom{\rule{0ex}{0ex}}\\ \text{samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% two-tailed confidence interval with}\\ \text{n-1 degrees of freedom (from Appendix A).}\end{array}$

and

(B) Any represented value of the annual fuel utilization efficiency or other measure of energy consumption of a basic model for which consumers would favor higher values shall be less than or equal to the lower of:

(1) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(2) The lower 97 1/2 percent confidence limit (LCL) of the true mean divided by 0.95, where:

$\begin{array}{c}LCL=\stackrel{—}{x}-{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the number of}\phantom{\rule{0ex}{0ex}}\\ \text{samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed confidence interval with}\\ \text{n-1 degrees of freedom (from Appendix A).}\end{array}$

(ii) For the lowest capacity basic model of a group of basic models of those sectional cast-iron boilers having identical intermediate sections and combustion chambers, a sample of sufficient size shall be randomly selected and tested to ensure that -

(A) Any represented value of estimated annual operating cost, energy consumption or other measure of energy consumption of a basic model for which consumers would favor lower values shall be greater than or equal to the higher of:

(1) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(2) The upper 97 1/2 percent confidence limit (UCL) of the true mean divided by 1.05, where:

$\begin{array}{c}UCL=\stackrel{—}{x}+{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the number of}\phantom{\rule{0ex}{0ex}}\\ \text{samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed confidence interval with}\\ \text{n-1 degrees of freedom (from Appendix A).}\end{array}$

and

(B) Any represented value of the fuel utilization efficiency or other measure of energy consumption of a basic model for which consumers would favor higher values shall be less than or equal to the lower of:

(1) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(2) The lower 97 1/2 percent confidence limit (LCL) of the true mean divided by 0.95, where:

$\begin{array}{c}LCL=\stackrel{—}{x}-{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the number of}\phantom{\rule{0ex}{0ex}}\\ \text{samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed confidence interval with}\\ \text{n-1 degrees of freedom (from Appendix A).}\end{array}$

(iii) For the highest capacity basic model of a group of basic models of those sectional cast-iron boilers having identical intermediate sections and combustion chambers, a sample of sufficient size shall be randomly selected and tested to ensure that -

(A) Any represented value of estimated annual operating cost, energy consumption or other measure of energy consumption of a basic model for which consumers would favor lower values shall be greater than or equal to the higher of:

(1) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(2) The upper 97 1/2 percent confidence limit (UCL) of the true mean divided by 1.05, where:

$\begin{array}{c}UCL=\stackrel{—}{x}+{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the number of}\phantom{\rule{0ex}{0ex}}\\ \text{samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed confidence interval with}\\ \text{n-1 degrees of freedom (from Appendix A).}\end{array}$

and

(B) Any represented value of the fuel utilization efficiency or other measure of energy consumption of a basic model for which consumers would favor higher values shall be less than or equal to the lower of:

(1) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(2) The lower 97 1/2 percent confidence limit (LCL) of the true mean divided by 0.95, where:

$\begin{array}{c}LCL=\stackrel{—}{x}-{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the number of}\phantom{\rule{0ex}{0ex}}\\ \text{samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed confidence interval with}\\ \text{n-1 degrees of freedom (from Appendix A).}\end{array}$

(iv) For each basic model or capacity other than the highest or lowest of the group of basic models of sectional cast-iron boilers having identical intermediate sections and combustion chambers, represented values of measures of energy consumption shall be determined by either -

(A) A linear interpolation of data obtained for the smallest and largest capacity units of the family, or

(B) Testing a sample of sufficient size to ensure that:

(1) Any represented value of estimated annual operating cost, energy consumption or other measure of energy consumption of a basic model for which consumers would favor lower values shall be greater than or equal to the higher of:

(i) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(ii) The upper 97 1/2 percent confidence limit (UCL) of the true mean divided by 1.05, where:

$\begin{array}{c}UCL=\stackrel{—}{x}+{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the}\phantom{\rule{0ex}{0ex}}\\ \text{number of samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed}\\ \text{confidence interval with n-1 degrees of freedom (from Appendix A).}\end{array}$

and

(2) Any represented value of the energy factor or other measure of energy consumption of a basic model for which consumers would favor higher values shall be less than or equal to the lower of:

(i) The mean of the sample, where:

$\begin{array}{c}\stackrel{—}{x}=\frac{1}{n}\sum _{i=1}^{n}{x}_{i}\\ \text{and,}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; n is the number of samples; and}\phantom{\rule{0ex}{0ex}}{x}_{i}\phantom{\rule{0ex}{0ex}}\text{is the}\phantom{\rule{0ex}{0ex}}{i}^{\mathrm{th}}\phantom{\rule{0ex}{0ex}}\text{sample;}\end{array}$

Or,

(ii) The lower 97 1/2 percent confidence limit (LCL) of the true mean divided by 0.95, where:

$\begin{array}{c}LCL=\stackrel{—}{x}-{t}_{.975}\left(\frac{s}{\sqrt{n}}\right)\\ \mathrm{And}\phantom{\rule{0ex}{0ex}}\stackrel{—}{x}\phantom{\rule{0ex}{0ex}}\text{is the sample mean; s is the sample standard deviation; n is the}\phantom{\rule{0ex}{0ex}}\\ \text{number of samples; and}\phantom{\rule{0ex}{0ex}}{t}_{0.975}\phantom{\rule{0ex}{0ex}}\text{is the t statistic for a 97.5% one-tailed}\\ \text{confidence interval with n-1 degrees of freedom (from Appendix A).}\end{array}$

(v) Whenever measures of energy consumption determined by linear interpolation do not agree with measures of energy consumption determined by actual testing, the values determined by testing must be used for certification.

(vi) In calculating the measures of energy consumption for each unit tested, use the design heating requirement corresponding to the mean of the capacities of the units of the sample.

(vii) Reported values. The represented value of annual fuel utilization efficiency must be truncated to the one-tenth of a percentage point.

(b) Certification reports.

(1) The requirements of § 429.12 are applicable to residential furnaces; and

(2) Pursuant to § 429.12(b)(13), a certification report shall include the following public product-specific information:

(i) Residential furnaces and boilers: The annual fuel utilization efficiency (AFUE) in percent (%) and the input capacity in British thermal units per hour (Btu/h).

(ii) For cast-iron sectional boilers: The type of ignition system for gas-fired steam and hot water boilers.

(3) Pursuant to § 429.12(b)(13), a certification report shall include the following additional product-specific information: For cast-iron sectional boilers: a declaration of whether certification is based on linear interpolation or testing. For hot water boilers, a declaration that the manufacturer has incorporated the applicable design requirements.

(4) For multi-position furnaces, the annual fuel utilization efficiency (AFUE) reported for each basic model must be based on testing in the least efficient configuration. Manufacturers may also report and make representations of additional AFUE values based on testing in other configurations.

[76 FR 12451, Mar. 7, 2011; 76 FR 24765, May 2, 2011, as amended at 76 FR 38292, June 30, 2011; 81 FR 2646, Jan. 15, 2016]