Statistical processing

Source data

The data basis for the 2023 survey is a web-based survey sent to a sample of 49,325 households, stratified by housing type: single-family houses (typically detached houses, terraced houses, etc.), apartments, and holiday homes used for year-round living. Single-family houses are further divided by primary heating source, i.e., whether they have district heating according to the BBR (Building and Housing Register). Dormitories and residential institutions are not included in the selection. Holiday homes without year-round residence are addressed by asking respondents who indicated that they own or have access to a holiday home to complete a questionnaire on heating sources and the use of firewood, etc.

Frequency of data collection

The statistics are compiled every other year.

Data collection

The questionnaire has been integrated into a digital platform/web survey, allowing for the addition of illustrations, photos, and detailed explanations of technologies, measurement units, and concepts. See the attached questionnaire for 2023 (heating season 1 June 2023 - 31 May 2024).

For the 2023 survey, 49,325 households were selected to participate. They were chosen from a population of 2,806,291 households. The criteria for inclusion in the population are that it must be a unique address in the Building and Housing Register (BBR) with at least one registered CPR number according to the CPR register. A unique address could be, for example, an apartment in a multi-story building. At least one CPR number means that there must be at least one person who has registered the address as their official residence. The sample is stratified by housing type: single-family houses (typically detached houses, terraced houses, etc.), apartments, and holiday homes used for year-round living. Single-family houses are further divided by primary heating source, i.e., whether they have district heating according to the BBR. Dormitories and residential institutions are not included in the selection. The stratification follows the distribution below.

Strata: - Houses with district heating 16,957 - Houses without district heating 23,085 - Apartments 5,674 - Occupied holiday homes 3,609

Total 49,325

Apartments are deliberately underrepresented in the sample because the incidence of wood-burning installations in apartments is low, but they occur more frequently in single-family houses or occupied holiday homes. For houses with district heating, a lower selection rate is used than for houses without district heating, as wood-burning installations are more common in the latter category.

Data collection was carried out from 23 May 2024 to 26 June 2024. The first invitation letter was sent from 23 to 27 May 2024 via digital mail (e-Boks or digital mail through the public sector). The survey was sent to 49,325 individuals, of which 3,271 did not receive it due to opting out of e-Boks or digital mail, or due to rejection upon loading. 46,054 households had the opportunity to respond to the survey, of which 18,990 chose to do so, giving a net response rate of 38.5%.

Data validation

In the questionnaire, there is an advanced validation of respondents' statements with maximum limits for different types of firewood measurements. The maximum values for consumption are set at 50 for firewood sold in cubic meter measures or wood towers, while for wood briquettes and wood pellets, an upper weight limit of 25,000 kg is set.

The questionnaire also includes an advanced validation of respondents' statements regarding the consumption of firewood, etc., with different maximum limits for different firewood measurements. The maximum values for consumption are set at 50 for firewood sold in cubic meter measures or wood towers, while for wood briquettes and wood pellets, an upper weight limit of 25,000 kg is set.

The response dataset has also been reviewed for various types of errors using both manual and automated methods. - Obvious errors and misunderstandings. - In some cases, residents in occupied holiday homes have erroneously reported the same consumption in both the dwelling (holiday homes for year-round living) and under holiday homes. The consumption in the holiday home is set to 0 to avoid double counting and potential overestimation of total firewood consumption. - In cases where "other" or "unknown" heating systems were selected as the primary heating source, but one or more heating sources were specified in the detailed text, the primary and possibly secondary heating sources have been recoded.

• Automatic corrections
• If a primary heating source based on burning wood, etc., is indicated, there must be a physical consumption of firewood, wood briquettes, or wood pellets. The same applies to any secondary heating sources unless it is stated that they were not in use during the season. Responses where no consumption quantities are indicated are imputed based on median values from the total energy consumption of other respondents, converted to equivalent cubic meters of sawn, split, and stacked wood. This type of firewood was chosen as it is the most common.
• Outliers at the top and bottom of firewood consumption have been identified and corrected. Data includes outliers with unrealistically high consumption as well as responses without reported consumption, even though this is unrealistic. In principle, excessively high values are imputed with the 97.5th percentile of comparable responses, while very low responses are imputed with the 2.5th percentile.

Data compilation

The response dataset is manually and automatically reviewed for obvious errors and misunderstandings, logical errors, and inconsistencies, as well as for missing or excessively high/low values. Missing values are imputed based on the other responses, and very high or low values are adjusted down or up based on the 97.5th or 2.5th percentiles, respectively. The adjusted data is corrected as described in section 3.6, and the results are then scaled to the population level. The scaling weights used are calculated to reflect all households in Denmark and result from stratification. In line with previous surveys, these scaling weights are also used in a corrected form to scale the results for unoccupied holiday homes; see section 3.06 for correction. Finally, the scaled and corrected data is tabulated.

Energy consumption is calculated as the calorific value of the consumed amounts of firewood, etc., using the following conversion factors:

• Cubic meters of firewood: 5.94 GJ per unit
• Cubic meters of sawn, split, and stacked firewood: 7.40 GJ per unit
• Box cubic meters of firewood: 4.39 GJ per unit
• Wood tower: 14.50 GJ per unit
• Wood briquettes: 17.50 GJ per 1,000 kg
• Wood pellets: 17.50 GJ per 1,000 kg

Adjustment

Three corrections have been made to the edited consumption data:

Correction for unreported firewood consumption in rented holiday homes: For holiday homes rented out from September 2023 to April 2024, where the reported consumption quantities do not cover tenants' or borrowers' consumption, this is corrected by multiplying the consumption by the ratio of the number of weeks rented out during the period to its total length. The correction factor is calculated as (Number of weeks rented/(Number of weeks from September to April - Number of weeks rented)). The respondent's reported consumption is corrected as follows:

Corrected consumption = Immediate response on consumption ∙ (1+correction factor).

This correction is specific to each relevant response, i.e., for holiday homes without permanent residents that were rented out from September to April and where the reported consumption does not cover the tenant's consumption of firewood, etc.

Degree day correction: The survey asks about firewood consumption, etc., during the period from June 2023 to May 2024. To convert consumption from the 2023/2024 heating season to the calendar year 2023, the consumption is corrected for temperature variations using the ratio between the number of degree days in the heating season and the calendar year. The Danish Meteorological Institute (DMI) publishes månedsvise opgørelser on the number of degree days, and for the calendar year 2023, the total number of degree days was 2754, while the number of degree days in the 2023/2024 heating season was 2696.

The correction factor is therefore calculated as 2754/2696=1.0218, meaning that the consumption calculated from the collected data is multiplied by this factor.

This correction applies to all calculated consumption quantities and derived energy calculations, but not the number of firewood installations.

Correction of scaling weights for unoccupied holiday homes: To scale the responses for unoccupied holiday homes, the scaling weights calculated for responses regarding dwellings are used. Without correction, these weights would lead to an overestimation of the number of unoccupied holiday homes in the survey compared to the official statistics from BBR and Statistics Denmark (BOL101)[www.Statbank.dk/BOL101]. Based on how many respondents in the survey state that they own/use a holiday home in Denmark, this corresponds to 360,572 unoccupied holiday homes in Denmark. The official figure for 2023 is 222,066, and the difference is partly due to several households jointly owning a holiday home. To provide a more accurate picture of unoccupied holiday homes, the scaling weights are adjusted by the ratio between the calculated number and the official figures.

The correction factor for holiday homes in 2023 is as follows: 360,572/222,066 = 1.62372.

All results concerning the number and total firewood consumption for unoccupied holiday homes are divided by a correction factor of 1.62372 (or multiplied by 0.6159) to avoid overestimation.