Annual Industrial Consumption of Energy Survey (ICE)

Detailed information for 2001

The survey provides estimates of energy consumption by manufacturing establishments in Canada.

Data release - August 20, 2002

• Questionnaire(s) and reporting guide(s)
• Description
• Data sources and methodology
• Data accuracy

Description

The survey, which is funded by Natural Resources Canada and Environment Canada, provides estimates of energy consumption by manufacturing establishments in Canada.

Federal (Natural Resources Canada, Environment Canada) and several provincial agencies are provided with data on a regular basis, as are several forest product associations.

Reference period: year

Collection period: mid-January to May

Subjects

• Energy
• Energy consumption and disposition

Data sources and methodology

Target population

The target population comprises manufacturing establishments in Canada, except the major energy consumers that are the target population of the Quarterly Industrial Consumption of Energy survey.

Instrument design

The questionnaire comprises consumption of electricity, natural gas, propane, diesel, other middle distillates, heavy fuel oil, wood, spent pulping liquor, refuse, and steam. The energy uses of interest are consumption as fuel, consumption to produce steam for sale, consumption in production of electricity, and consumption for non-fuel use.

The questionnaire is respondent completed.

Three versions of the questionnaire are used: a general version, a version for the petroleum products industry (North American Industrial Classification System - NAICS 324110, 211114), and a Regional Office version.

Sampling

This is a sample survey with a cross-sectional design.

The sample design encompasses both the Quarterly Industrial Consumption of Energy survey (Survey ID 2166) and the Annual Industrial Consumption of Energy survey.

The ICE survey population is the population defined by the Annual Survey of Manufacturing (ASM), which is all manufacturing establishments contained on Statistics Canada's Business Register. The ASM applies take-none thresholds to minimize the collection of data from smaller establishments and selects a sample from the remaining establishments. The ICE survey population excludes an additional 5% of the establishments in terms of shipments to cover 95% of the total manufacturing shipment value for the ASM sample.

The frame is divided into industry groups based on estimation requirements. The shipment value from the ASM principal statistics file is found to be reasonably correlated with the overall energy consumption variable on the same file, and is used as a stratification variable. The population is divided into 3 strata based on the Lavalée-Hidiroglou procedure and the sample is allocated to these strata to give a Coefficient of Variation (CV) of 5%.

To reduce sampling variability, the largest size stratum, within each industry grouping, is sampled with certainty. This stratum is called a "take-all" stratum. If there are other strata created within an industry grouping, establishments are sampled by a probability mechanism. These strata are called "take-some" strata.

The largest energy consumers that have data collected through the Quarterly Industrial Consumption of Energy Survey are included in the appropriate take-all strata. Additionally, if the number of establishments within a particular industry grouping is less than or equal to 50, all units are included in the sample (i.e. these industry grouping would have just 1 take-all stratum.)

For the take-some strata where sampling is done, Statistics Canada's Generalized Sampling System (GSAM) is used to randomly select the units for which data is to be collected. All sampled units are assigned a sampling weight. The sampling weight is a factor attached to each sampled unit to indicate how many similar units the selected unit represents in the population. This weight allows estimates for the population to be obtained from the sample.

Data sources

Responding to this survey is mandatory.

Data are collected directly from survey respondents.

The collection period begins with the mailing of the questionnaires to the selected establishments in mid-January. The annual survey data for the larger energy consumers is collected by Statistics Canada's Head Office, and the data for smaller energy consumers is collected by a Statistics Canada Regional Office.

Phone and fax follow-ups begin in March for establishments that have not yet responded. The Regional Office collects data from late responders using Computer Assisted Telephone Interviews.

The collection period ends in May.

View the Questionnaire(s) and reporting guide(s).

Error detection

The following methods of error detection are used:

. Edits are performed during capture to ensure that keying errors are corrected. Other consistency edits are also done once data capture is complete to ensure that related data values are consistent with each other.

. Statistical editing is used to compare reported values to a distribution of data values. If the reported values are at either extreme end of the distribution, the respondent who reports such data fails the edit and becomes a candidate for possible follow-up.

. For units that have no previous data values to compare with, the most recent ASM fuel and electricity consumption values are used as a comparison, where ASM data definitions are similar to ICE data definitions.

. The relative percentage difference between the ICE data and ASM data is compared to the distribution of relative percentage changes over all sampled units. A similar distribution is created for joule based total energy consumption. If a unit fails both types of edits, the unit becomes a priority for edit failure phone follow-up. The unit has a lower priority for follow-up if just one type of edit fails.

Any edit failures not resolved are flagged in order to be excluded from use in imputation.

Imputation

Statistical imputation is used for cases of total non-response. A trend factor is calculated for industry groupings and applied to the last period's response for respondents with missing data in that grouping to obtain an imputed response for the current period.

Estimation

A single estimation process is used for the manufacturing sector of the Quarterly and Annual Industrial Consumption of Energy surveys. The Generalized Estimation System is applied to calculate energy consumption estimates for all 3 digit and selected 4, 5 and 6 digit NAICS manufacturing industries.

The estimation strategy takes into consideration a number of factors: that ICE is a sub-sample of the ASM, that only 95% of the ASM is covered, and that the ASM also uses take-non boundaries.

The estimates are calibrated to the most recent shipment values available from the ASM. This corrects for sampling errors and is an uncovered portion adjustment.

Quality evaluation

Fuel and electricity consumption estimates are compared with those from the Annual Survey of Manufactures and from energy surveys such as the Electric Power Thermal Generating Stations Fuel Consumption survey.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.

Confidentiality analysis includes the detection of possible direct disclosure, which occurs when the value in a tabulation cell is composed of a few respondents or when the cell is dominated by a few companies.

Data accuracy

The response rate to the survey is approximately 90%. As a consequence, minimal bias resulting from non-response is introduced.

Since this survey was based on probability sampling the potential for error caused by sampling can be measured. A standard measure of sampling error is the coefficient of variation (CV). Estimates with smaller CVs are more reliable than estimates with larger CVs. At the Canada level, the target CV is between 1 and 2%, 0% for industry groupings that are a census of the target population, and less than 20% for the remaining industry groupings.