View all publications

Annex 2: Methodology and data for estimating emissions from fossil fuel combustion

New Zealand emission factors are based on GCV since energy use in NZ is conventionally reported in gross terms, with some minor exceptions. The convention adopted by New Zealand to convert GCV to NCV is to follow the OECD/IEA assumptions that:

  • NCV = 0.95*GCV for coal and liquid fuels
  • NCV = 0.90*GCV for gas

A2.1 Emissions from liquid fuels

Activity data: Statistics New Zealand conducts the Delivery of Petroleum Fuels by Industry survey. Statistics New Zealand is New Zealand's national statistical office and administers the Statistics Act 1975.

The Delivery of Petroleum Fuels by Industry is a quarterly survey. The purpose of the survey is to provide data on the amount of fuel delivered by oil companies to end-users and other distribution outlets. Each oil company in New Zealand supplies Statistics New Zealand with the volume of petroleum fuels delivered to resellers and industry groups. It is assumed there is a 5% uncertainty associated with the sectoral energy allocation although the annual totals are likely to be more certain (MED, 2003).

The survey is a census which means there is no sampling error. The main sources of non-sample error are:

  • Respondent error: Statistics New Zealand makes every effort to confirm figures supplied by the respondents, and given assurances of accuracy, Statistics New Zealand are bound to accept them. If a discrepancy is discovered at a later date, revised figures are supplied at the earliest possible opportunity.
  • Processing error: there is always the possibility of error. However, Statistics New Zealand has thorough checking procedures to ensure that the risk of processing errors is minimised.

Emission factors: CO2 emission factors are described in Table A2.1. The CO2 emission factors for oil products are from the New Zealand Refining Company (NZRC) data, import data from industry and from a New Zealand source (Baines, 1993 [The LPG CO2emissions factor was confirmed by checks of 2102 gas data.] ). The same values are used for each year of inventory. There is a direct relationship between fuels' carbon content and the corresponding CO2 emissions during combustion. However, the carbon composition of oil products is not closely monitored and there will be variation over time, depending on the crude oil used in production. The NZRC estimates the uncertainty in emission factors to be within 5% (MED, 2003).

New Zealand's review of emission factors (Hale and Twomey, 2003) identified a number of non-CO2 emission factors (Tables A2.2 and A2.3) where the supporting information (Bone et al., 1993; Waring et al., 1991) was assessed to be insufficient to retain the country-specific emission factors used in previous inventories. The changes mainly affected the mobile combustion category. Where a country-specific value was not available, the emissions factor used by New Zealand is either the IPCC value that was the closest match to New Zealand's conditions or the mid-point value from the IPCC range.

Many of the sources in the stationary combustion category already used the IPCC (1996) emission factors, however there were minor changes to the 2002 inventory related to the re-interpretation of the IPCC tables resulting from the Hale and Twomey review. The changes were made to obtain the closest match to New Zealand conditions. The decision to change from the country-specific emission factors to IPCC values was confirmed by an expert review panel prior to including the emission factors in the inventory. There was no further review of emission factors for the 2003 inventory. All emission factors from the IPCC Guidelines are converted from NCV to GCV.

Table A2.1 CO2 emission factors used in the energy sector

View CO2 emission factors used in the energy sector (large table).

Table A2.2 CH4 emission factors used in the energy sector

View CH4 emission factors used in the energy sector (large table)

Table A2.2 N2O emission factors used in the energy sector

View N2O emission factors used in the energy sector (large table)

A2.2 Emissions from solid fuels

Activity data: Statistics New Zealand Coal Sales survey is an ongoing quarterly survey which commenced on 1 March 1981. The purpose of this survey is to measure the amount of coal which is sold and available to users. The target population is all coal mines and major resellers of coal in New Zealand. Completion of the survey has been approved by the Minister of Statistics and returning the completed questionnaire, duly signed, is a compulsory requirement under the Statistics Act 1975.

The survey is a full coverage of the sector and therefore there are no sampling errors. Non-sampling errors in the survey data may result from errors in the sample frame (i.e. units with the wrong New Zealand Standard Industrial Classification), respondent error (i.e. wrong values supplied), mistakes made during processing survey results or non-response imputation. Statistics New Zealand adopts procedures to detect and minimise these types of errors but they may still occur and they are not quantifiable.

The three ranks of coal measured are bituminous, sub-bituminous and lignite coal. From 1988 onwards, the coal sales questionnaire broke coal sales into seven end use sectors, however these do not match the IPCC sectors. The sectoral shares of coal use that can be used for the UNFCCC inventory are based on Coal Research Limited's (CRL) survey of sectoral coal use for 1990 and 1995. Data are interpolated between 1990 and 1995 and extrapolated for all years beyond 1995. The exceptions are for the coal used for steel manufacture, electricity production and the residential household sector where the MED use figures from Statistics New Zealand's Coal Sales survey. Sectoral shares are calculated by:

  • Summing the four calendar year quarters of coal sales data from Statistics New Zealand's Coal Sales survey.
  • Subtracting coal exports and coal used by the residential sector (from Statistics New Zealand's Coal Sales survey) and coal used for steel and electricity (both known accurately).
  • Dividing CRL annual coal tonnage for each sector by the total (excluding exports, steel, electricity and residential coal use) to give sectoral shares of coal use for 1990 and 1995.
  • Interpolating sectoral shares between 1990 and 1995 and extrapolating for beyond 1995.
  • Applying the sectoral share estimates to the Coal Sales survey's total coal sold (excluding exports, electricity, steel and residential coal use).

The process of dividing coal use between different sectors will introduce uncertainty larger than the uncertainty in total coal sales. Uncertainty is also introduced from the assumption that coal used by sector is an average of the different ranks. An uncertainty of ± 5% is assumed to cover these issues (MED, 2003).

The sectoral partitioning used for coal was examined in 2003 by MfE officials. There was concern in extrapolating sectoral allocations from 1995 to 2002 given some probable changes in sectoral coal usage. However, coal industry experts (W.Hennessy pers comm.) did not consider a survey could be justified because of the difficulty and expense in collating and verifying data from a number of sectors. Furthermore, data for the major categories of coal exports, coal used by the residential sector, and coal used for steel and electricity are all known accurately and are not affected by the sectoral partitioning.

Emission factors: The value for sub-bituminous coal (91.2 kt CO2/PJ) is used to calculate New Zealand's emissions from coal burning (Table A2.1). Using only the sub-bituminous value for all ranks of coal is a reasonable assumption for New Zealand as the bulk of the high quality bituminous coal is exported and all coal used in public electricity generation is of sub-bituminous rank (MED, 2003). The range in emission factors across all grades of coal is 5.5%. Therefore the estimated uncertainty in coal emission factors is taken as ± 3% (MED, 2003). An uncertainty of ± 2% is used for the sub-bituminous coal used in public electricity generation. All New Zealand values are within 2% of the IPCC defaults (1996). The non-CO2 emission factors are shown in Tables A2.2 and A2.3.

A2.3 Emissions from gaseous fuels

Activity data: the Natural Gas Corporation (NGC) has contracts with large users that allow metering errors of ± 2%. Whenever the error between the meter reading and actual gas supplied exceeds 2%, adjustments are made to the reported quantities of gas supplied. The uncertainty is therefore assumed to have an upper limit of ± 2% (MED, 2003).

Emission factors: the emission factors for natural gas used in distribution and sold to large users are shown in Table A2.1. The values are calculated by averaging daily gas composition data supplied by industry. The composition, hence the emissions factor, varies slightly between daily measurements. Taking annual bounds, it is estimated that the uncertainty in the natural gas emission factors is less than 1.7% (MED, 2003).

It has been assumed that half of the gas in the system is from the Maui gas field with an average CO2 emissions factor of 52.0 kt CO2/PJ in 2002 and the other half is treated gas (52.6 kt CO2/PJ). The average value of 52.3 kt CO2/PJ is used for the 2003 inventory (Table A2.4).

Table A2.4 Variation in CO2 emission factors for natural gas

Year Maui (kt CO2 / PJ) Treated (kt CO2 / PJ) Average (kt CO2 / PJ)

1990

53.2

52.4

52.8

1991

52.9

52.8

52.8

1992

52.9

52.7

52.8

1993

52.6

52.5

52.5

1994

52.4

52.2

52.3

1995

52.1

52.9

52.5

1996

52.2

52.9

52.6

1997

52.3

52.4

52.4

1998

52.1

52.2

52.1

1999

51.8

52.4

52.1

2000

52.1

52.1

52.1

2001

51.9

52.6

52.3

2002

52.3

52.5

52.4

2003

52.0

52.6

52.3

A2.4 A Tier 2 assessment of road transport emissions

In line with good practice, New Zealand has developed a Tier 2 model for estimating CO2, CH4 and N2O emissions from mobile combustion to support the Tier 1 estimate. The Tier 2 model is described below.

In New Zealand every vehicle is registered, and licensed annually, with information held on the make, model and other attributes of the vehicle in a centralised database - the Motor Vehicle Register. This provides information on the composition of the vehicle fleet and variation over time. The New Zealand Ministry of Transport maintains a Vehicle Fleet Model (VFM) which translates the fleet composition profiles into environmental performance measures. Within this process, the VFM calculates fuel used by the vehicle fleet, based on data as available and supporting estimates on vehicle fuel economy and usage.

The inputs to the model, which strongly influence how much fuel is used, include:

  1. Fuel economy estimates for the different vehicle classes i.e. technical efficiency, type of engine and emission control technology, and vehicle/engine size. Also, for the heavy freight sector, consideration of the load transported.
  2. Estimates of activity - distance travelled annually by vehicle class and also the driving conditions. This involves distributing the Vehicle Kilometres Travelled (VKT) over different road types and congestion levels. Standard conventions, common to road and traffic engineering, are used.

Limited data sources are available for direct use in calibrating this model. Determining the input factors requires an expert judgement of likely efficiency of the fleet and driving patterns. These factors will be refined as data sources develop. For example, new entrants to the fleet now have their standard reference fuel consumption measure recorded in the database, and more sophisticated procedures are in process for monitoring VKT. In time, this information will be available as a routine, coherent data source. In the meantime, an expert assessment of fuel economy is used. Numbers are based on knowledge of the available vehicle technology and fuel consumption performance relationships, and variability in fuel use with different driving conditions.

The VKT estimates are derived from comparison with various survey sources and derived estimates, including local and national road surveys and analysis of typical annual mileages by vehicle type, using information from the Motor Vehicle Registry. Then, an expert judgement is made as to how this total is allocated to vehicle class.

The fleet model is built up from a series of data spread sheets. The major steps are:

  • The base data is the fleet for each calendar year split by vehicle class (e.g. cars, buses, light commercial vehicles, heavy goods vehicles by weight categories, and motorcycles) and also by fuel (primarily diesel and petrol).
  • For each vehicle class, an assumption on the average fuel economy for that class is added. This assumption is based on known data on vehicle age, makes and models and is weighted by the percentage of different vehicles in that class. For instance, New Zealand has a large market for pre-used vehicle imports from Japan in addition to Australian, US and EURO standard vehicles.
  • The fleet model then multiplies the likely fuel economy for each class, by a figure for annual VKT by each class, across the range of different driving conditions.
  • The model output is in fuel consumed in litres, per km in different driving conditions, and then the total per annum for the vehicle classes in the fleet. Finally, using country specific (for CO2) and IPCC (for CH4 and N2O) emission factors for the different fuels (petrol, diesel, LPG and CNG), greenhouse gas emissions are calculated from the national fleet per annum.

Because there are a number of inputs to the model that are based on expert judgement there is need for sensitivity analysis. In particular, the input figures of indicative vehicle fuel economy for each vehicle class and distance kilometres travelled for each class are varied to assess the sensitivity of the model to changes in those variables.

The focus of the model has been on refining understanding of the light fleet, rather than the heavy fleet. The fuel economy of the heavy vehicle classes (which use the majority of the diesel fuel in New Zealand) is particularly sensitive to changes in load carried and terrain, both of which vary significantly in New Zealand. In time, more data are expected to become available on the characteristics of heavy vehicle travel, allowing refinements to be made to the model. The model is based on a micro-inventory approach which identifies the factors that most affect the confidence in the results. It is also noted that off road, non-registered vehicles, that commonly use diesel, are not included in the model.

Emission factors: The model uses the pre-2002 inventory petrol and diesel emission factors, as the model in its present form cannot split the petrol use into regular and premium as per the 2003 inventory. These emission factors are:

  • CO2: 66.6 kt/PJ for petrol and 68.7 kt/PJ for diesel
  • Petrol: density 0.75, GCV 45.9 MJ/kg; gives CO2 emissions at 2.30 kg/litre
  • Diesel: density 0.84, GCV 45.7 MJ/kg; gives CO2 emissions at 2.64 kg/litre

The 1996 IPCC guidelines provided the emission factors for CH4 and N2O. They were selected as per the corresponding vehicle technology categories used in the vehicle fleet model. The calculation process starts from the g/kg of fuel consumption indices. These are translated to the required g/km emission rates by correlating with the fuel consumption rates for New Zealand fleet type vehicles, to calculate total output against the fleet VKT baselines.