2 Methodology

The method recommended by the SWAP for measuring the composition of waste is based on sorting and weighing sub-samples of the waste stream. Waste Not Consulting uses a visual classification method for these solid waste audits. These two methods of waste classification are compared below

2.1 Comparison of methods of waste classification

The SWAP includes two separate procedures for measuring the composition of solid waste: Procedure One for the classification of domestic waste (primarily bagged refuse, as is commonly collected by local territorial authorities) and Procedure Two for the classification of waste entering disposal facilities.

For both procedures, the recommended method for measuring the composition of waste is based on sorting and weighing statistically accurate sub-samples of the waste stream. For Procedure Two a visual classification method is also presented, “whereby an observer estimates the proportions of refuse belonging to the primary waste classifications”.³ However, the visual classification method is not recommended by the SWAP because “visual classification potentially introduces a bias into the survey results, as the data is dependent on observation rather than measurement”.⁴

Since 2003, Waste Not Consulting has been using visual classification techniques in preference to the sort-and-weigh methodology recommended for SWAP Procedure Two audits of disposal facilities. Waste Not Consulting used the sort-and-weigh methodology for several disposal facility audits in the period 2001/03, but, having encountered several difficulties with this methodology, changed to visual survey techniques.

The primary driver for this change was occupational safety and health. During a sort-and-weigh audit the physical handling of waste of unknown and potentially hazardous composition is unavoidable, and this handling involves risks that must be reasonably mitigated to ensure the safety of workers. For some types of waste audits, such as those involving office or bagged domestic waste (eg, those using Procedure One of the SWAP) in controlled environments, Waste Not considers that the risks of physically handling waste can be adequately mitigated and continues to use sort-and-weigh techniques.

In other environments, however, such as in large landfills and transfer stations, Waste Not does not consider that the risks of physically handling waste for a sort-and-weigh audit can be reasonably mitigated. This is due to both the physical environment and the waste material itself. The physical environment in these facilities is hazardous for sort-and-weigh audits, and commonly involves heavy vehicle traffic, uncovered work areas, difficult conditions underfoot (such as mud or compactor tracks) and hazardous dust (such as from asbestos disposal). The ‘covered and paved’ sorting area recommended by the SWAP⁵ for disposal facility audits is rarely available at large disposal facilities.

Although the hazards of the physical environment for sort-and-weigh waste auditing at a landfill could potentially − with sufficient expenditure of resources − be lessened, the risks represented by handling the waste materials themselves can not, in Waste Not’s opinion, be reduced sufficiently. Landfills in industrialised urban centres, in particular, receive a wide range of waste materials, many of which are too hazardous to handle and sort manually. Dangerous materials include not only large and heavy items that can not be sorted by hand, such as concrete and steel from demolition activity, but also hazardous materials such as autoclaved medical waste and vehicle floc.⁶ In Waste Not’s opinion, there is no practical and safe method to sort and weigh many types of waste.

Although the health and safety considerations are paramount, there are other advantages to the visual surveying methodology. The primary one is the sample size of a visual survey compared to a sort-and-weigh survey. In Waste Not’s experience, one individual conducting a visual survey can generate data on over 300 vehicles in a single day. This compares to a sort-and-weigh audit of fewer than 100 vehicles per week, which requires a team of four people.⁷

With its much larger sample size, a visual survey can be used to generate stratified data with better precision levels than a sort-and-weigh audit. The visual surveys undertaken for the Waste Data Programme, for example, categorise waste loads into seven different activity source types (see Section 2.4). This compares to the two types of sources (business and residential) recommended by the SWAP. With a sample size of fewer than 100 vehicles, a sort-and-weigh audit would be unlikely to produce precise or accurate results for any greater number of source classifications.

The large sample size is a particular advantage for disposal facilities that have both a public-access transfer station and a discharge face (tip face) for commercial loads, such as Silverstream Landfill in Hutt City and Green Island Landfill in Dunedin. The previous 2003/04 SWAP audits at these sites combined the transfer station and landfill waste streams for analytical purposes. Using visual survey techniques in the current programme, analysis of the two waste streams can be provided separately.

It is acknowledged that the visual classification of the composition of waste loads is inherently more prone to bias than the sorting and weighing of an entire load of refuse, which the SWAP recommends for cars and trailers. It is uncertain, however, whether visual estimation of a large number of truck loads produces results that are more or less accurate than sorting and weighing 140 kilogram subsamples of a smaller number of truck loads of waste, as is recommended by the SWAP (Section 5.4.6). This subsample size of 140 kg was initially recommended by the SWAP because it was considered appropriate for the vehicle fleet of the time. No further studies are known to have been done to assess the suitability of the technique for the much larger vehicles currently in use; for example, the front-loaders using Silverstream Landfill average over 6 tonnes per load.

For reasons of both health and safety and practicality, Waste Not has never used the subsampling method recommended in Section 5.4.6 of the SWAP, in which the waste load is spread out and 20 kg samples taken from randomly chosen points. This method requires workers to either walk across the spread-out waste load or to lay out duckboards to walk on. The method also requires the constant availability of heavy equipment to spread and remove the waste, which increases the cost of the audit.

While, in theory, a random subsampling from trucks for a sort-and-weigh audit should produce accurate data, in Waste Not’s experience the physical separation of a subsample from a truckload of waste introduces bias into the auditing process. For example, a load of demolition material might contain a mixture of bulky, unwieldy items such as wooden beams, and fine debris such as concrete rubble. Waste Not has not found a satisfactory, practical method for obtaining a random subsample of such a load without introducing bias.

The SWAP recommends that surveyors undertaking visual classification of waste should calibrate their estimations by visually assessing a load before it being sorted and weighed and comparing the results.⁸ Although this concept may have value for assessing small waste loads, such as those that predominated during the 1992 WAP pilot trials, it is impractical for large loads, such as a six-tonne load in a front-loader truck. A high proportion of the waste entering a large disposal facility arrives in large truck loads, and typically less than 20 per cent arrives in cars and trailers. Given the safety and practical issues of sort-and-weigh measurements for trucks, the concept of calibrating for visual surveyors is of little relevance in large disposal facilities.

For visual surveying Waste Not Consulting relies on surveyors who have all gained experience from a range of sort-and-weigh audits. This experience calibrates their judgement by providing them with the knowledge of waste types and comparative material densities that is needed for visual surveying. For the 2007/08 Baseline Programme all four surveyors took part in a training session in which waste loads were assessed independently and the results compared.

The comparative cost of visual surveys and sort-and-weigh audits is the final factor in Waste Not’s continuing reliance on visual survey techniques. In a landfill situation, a sort-and-weigh audit can involve significant costs for the physical preparation of a suitable working area, rental of a marquee or similar shelter for auditing, provision of machinery and operators for subsampling waste loads and disposing of sorted waste, the hiring of casual staff for the sorting and weighing process, and the supply of appropriate health and safety equipment/treatments (eg, gas/dust masks, immunisations). In Waste Not’s experience, this extra expenditure, when necessary, can result in a sort-and-weigh audit costing twice as much as a visual survey. Given the health and safety concerns related to sort-and-weigh audits, along with the smaller data set and the uncertainty over the difference in the accuracy of the data from sort-and-weigh and visual audits, Waste Not can see no justification for the extra expense.

2.2 Visual waste survey

Using the technique developed by Waste Not for visual waste classification, as each vehicle was being unloaded at the transfer station or tip face the surveyor assessed the relative weight of each constituent present in the load on the basis of volume and density, in line with the secondary classifications given in Appendix 2. Absolute weights were not estimated; rather, the proportion of weight represented by each material was estimated. This data was recorded as a proportion, by weight, for each constituent present in the load.

For vehicle loads in which it was difficult to distinguish the individual constituents, a generic composition, based on previous sort-and-weigh surveys of that type of vehicle load, was used as a template for the composition, and was adjusted according to the materials that were visible. For example, a front-loader carrying large amounts of supermarket or restaurant waste was assessed as having a higher-than-average proportion of food waste.

For some specific waste streams, such as domestic bagged or wheelie bin refuse, an assumed composition was used for calculations, but no composition data was recorded for individual vehicle loads. Waste Not has collected an extensive range of data on this type of domestic refuse, and has tailored the composition data to suit the types of services (such as kerbside recycling) available to householders.

When the visual survey was completed, the data on the proportion of weights was combined with weighbridge records of the weight for each load, and a weight for each of the individual materials in the load was calculated. For small loads that were not weighed at the weighbridge the load weight was estimated based on previously determined averages for the specific vehicle and load type. These known averages have been made available to Waste Not by several disposal facilities that weigh every vehicle load entering the facility.

2.3 Waste classifications

A list of the classifications used for all of the Waste Data Programme audits is given in Appendix 2. There are 12 primary categories (eg, paper, plastic) and 22 secondary categories (eg, recyclable, multi-material).

2.4 Activity source of waste loads entering disposal facilities

For visual waste surveys such as those described in this report, Waste Not has developed a set of categories for the activity source of the waste. These are aimed at providing the information that is most useful for monitoring waste streams and effectively targeting waste minimisation initiatives and policies. The categories are more specific than the ‘residential’ and ‘business’ categories recommended by the SWAP. Information on the activity source of individual waste loads was gathered as the waste was unloaded, either by observation of the waste itself or by questioning the vehicle driver.

The categories used for all of the 2007/08 Waste Data Programme audits were:

1. construction and demolition (C&D) – waste materials from the construction or demolition of a building or structure (see Section 2.6 for a further discussion of C&D waste)
2. industrial/commercial/institutional (ICI) – waste from industrial, commercial and institutional sources (in practice it is not possible to gain more specific information on the waste source from truck drivers)
3. landscaping and earthworks – waste from landscaping activity and garden maintenance, both domestic and commercial, and from earthworks activity
4. residential – all waste originating from residential premises other than that covered by one of the other, more specific categories (includes drop-offs of bagged domestic waste)
5. kerbside collection – domestic-type waste collected from residential and commercial premises by either council or private kerbside waste collections
6. transfer station (for waste entering landfills) − consolidated loads transported from other transfer stations in the area
7. special waste (for waste entering landfills) − a subjective category that includes any substantial waste stream (such as biosolids, landfill cover material, infrastructural cleanfill, or industrial waste) that significantly affects the overall composition of the waste stream and may be markedly different from waste streams at other disposal facilities.

2.5 Identification of vehicle types

Procedure Two of the SWAP recommends separate sub-sampling of cars, trailers/utilities and trucks for disposal facility audits. Because visual surveying is capable of surveying all vehicles entering a facility, this subsampling was not required. However, loads carried by different vehicle types are not affected in similar ways by waste reduction initiatives (eg, waste in certain vehicles can more easily be separated into recyclable and/or compostable fractions), so all vehicles carrying refuse were identified by the surveyor as being one of the following vehicle types:

• cars

• trailers (including utes, vans and small trucks)

• front-loader trucks

• gantry trucks⁹

• compactor trucks

• huka trucks (including ‘smart skip’-type vehicles)

• other trucks (such as tip trucks and box trucks).

2.6 Taxonomy of C&D waste materials

The New Zealand Waste Strategy sets targets for waste minimisation, including a specific target for construction and demolition (C&D) waste. ‘Construction and demolition waste’, however, is not defined in the Strategy.

In previous work done for the Ministry for the Environment, Waste Not Consulting used a definition developed for the Construction and Demolition Waste Reduction Project, which was funded through the Sustainable Management Fund. One of the outputs of this project, Regulating Waste Management Under the Local Government Act 1974 & 2002: Guidance Notes, defines ‘construction and demolition waste’ as “Waste materials from the construction or demolition of a building, including the preparation and/or clearance of the property or site”.

The Guidance Notes definition would therefore exclude both man-made materials (such as concrete and asphalt) and natural materials (such as clay, soil and rock) when those materials are associated with infrastructure projects (such as road construction and maintenance), but would include those materials when they are generated by building-related projects. As a result, there is no way of assessing whether or not a load of clay, soil or rock is construction or demolition waste, other than by asking the vehicle driver. This is not possible while surveying in many instances, because cleanfill and cover material are not disposed of at the same location as general waste.

In the absence of a clear definition, and because of the difficulties involved in gathering information on the source of cover material, these materials are excluded from the analysis of C&D waste in the Waste Data Programme. C&D waste − or, in terms of the activity source of waste, waste resulting from construction and demolition activity − is taken to mean waste materials related to the construction or demolition of a building or structure, excluding natural materials such as clay or rock, unless they are disposed of in conjunction with man-made materials such as timber or concrete. Only fixed elements of a building or structure are considered C&D waste. Fixed elements of a house would include plasterboard and kitchen bench units, but would not include furniture and carpets. Waste from site preparation for a building or structure, such as soil and green waste, would be classified as originating from landscaping and earthworks activity.

Although these distinctions may seem arbitrary, they have proven to be functional while surveying. For example, soil that is disposed of at landfills is commonly associated with several different other waste materials, and could therefore be classified as resulting from one of three activities. Soil on its own, or with rocks or green waste, would likely be the result of landscaping and earthworks activity. Soil and timber would be classified as arising from construction and demolition activity, and soil and asphalt is likely to be from infrastructural development and would be classified as a special waste.

2.7 Health and safety

The surveyors undertaking the audits were equipped with all the personal protective equipment necessary for working in a disposal facility environment. Before the start of the survey the surveyors discussed safety procedures with site management.