2 The study context: previous dioxin data and reservoir estimates

2.1 Dioxin concentrations

The interpretation of the study’s results requires an understanding of the measurement of dioxins in terms of toxic equivalents. Following is a brief explanation of toxic equivalents and the analytical methods used.

Structure

Dioxin is the generic name for two groups of aromatic compounds with very similar molecular structure: the polychlorinated dibenzo-p-dioxins (PCDDs or dioxins) and the polychlorinated dibenzofurans (PCDFs or furans). Both groups of compounds can have up to eight chlorine atoms attached at different carbon atoms, and each individual compound is referred to as a congener. Each specific congener is identified by the number and position of chlorine atoms around the aromatic nucleus. In total, there are 75 possible PCDD congeners and 135 possible PCDF congeners. Congeners with the same number of chlorine atoms are known as homologues.

Toxicity

Congeners containing 1, 2 or 3 chlorine atoms are thought to be of no toxicological significance. However, the 17 congeners with chlorine atoms substituted in the 2,3,7,8-positions are thought to pose a risk to human and environmental health. Toxic responses include dermal toxicity, immunotoxicity, carcinogenicity, and adverse effects on reproduction, development and endocrine functions. Of the 17 congeners, the most toxic, and widely studied, congener is 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD). Increasing substitution from 4 to 8 chlorine atoms generally results in a marked decrease in toxicity.

In environmental media, the PCDDs and PCDFs occur as complex mixtures of congeners. To enable a complicated set of analytical results to be reduced to a single number, a system of toxic equivalent factors (TEFs) has been developed. The toxic equivalents method generates a set of weighting factors, each of which expresses the toxicity of a particular PCDD or PCDF congener in terms of an equivalent amount of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD). Multiplication of the concentration of a PCDD or PCDF congener by this TEF gives a corresponding 2,3,7,8-TCDD TEQ concentration. The toxicity of any mixture of PCDDs and PCDFs, expressed as 2,3,7,8-TCDD, is derived by summing the individual TEQ concentrations. This is reported as the ‘total TEQ’ for a mixture.

Although a number of toxic equivalents schemes have been developed, the most widely used is the International Toxic Equivalents Factor (I-TEF) scheme (Kutz et al., 1990). The I-TEF scheme was revised and expanded through the auspices of the World Health Organisation (WHO) by Van den Berg et al. (1998). More recently the WHO TEFs has been amended (van den Berg et al., 2006. This report, however, has not been amended to incorporate these most recent changes because the changes are minor, and the consequential and net effect on the overall TEQ calculation is also thought to be minor. Results of the site investigations in this study are reported using both the I-TEF and initial WHO (1998) TEF schemes to allow comparison with results from previous investigations using the I-TEF scheme and for future studies where the Ministry for the Environment has indicated they favour the use of the WHO TEF scheme.

OCDD screen vs full congener analysis

Because full congener dioxin analysis is expensive, a lower-cost analytical testing method was developed for determining heptachloro and octachloro dioxins and furans (ESR, 1992). Octachlorodibenzo-p-dioxin (OCDD) is the predominant dioxin congener contaminant in the PCP formulations used in New Zealand, and the hepta and octa congeners generally make up more than 95% of congeners found in contaminated soils on sawmill sites. Most results from previous investigations are for full congener analysis, and this should be noted when comparing these with the results from this study. In the calculations of the national reservoir, the ‘total TEQ’ values for the study results have been adjusted to take into account the difference between screen and full congener analysis.

The OCDD screen was based on USEPA Method 8290, and reproduced in Appendix J, Analytical Methods, in Pentachlorophenol Risk Assessment Pilot Study (CMPS&F, 1992). However, the OCDD screen now used by Agriquality laboratory, Gracefield, is consistent with USEPA Method 1613.

2.2 Previous dioxin data and reservoir estimates

The data presented in this section are taken from the New Zealand Inventory of Dioxin Emissions to Air, Land and Water, and Reservoir Sources (Ministry for the Environment 2000). The available data on dioxin levels in soils at sites where NaPCP or PCP was used is presented in Table 2.1 and is summarised as follows:

NaPCP use

• Small users (less than 20 tonnes used in total) - very limited dioxin concentration data (see Table 2.1) were available for small NaPCP users (all were believed to have used less than 4 tonnes) in the Canterbury region. The data were from samples taken from the top 200 mm of soil at a limited number of locations (CMPS&F, 1995).

• Medium users (20–100 tonnes used in total) - no data were available.

• Large users (more than 100 tonnes used in total) - no data were available.

• A very large user (about 1000 tonnes in total) - comprehensive data were available for the Waipa site (which was also a very large user of PCP in oil). This information is also provided in Table 2.1.

PCP in oil use

• A large user (about 420 tonnes in total) - some data were available covering a number of different sampling locations for one PCP-in-oil operation (Hanmer Springs). The mean value is given in Table 2.1.

• A very large user (about 2000 tonnes in total) - comprehensive data were available for the Waipa site.

The estimates of areas at the sites potentially affected by dioxin residues are also provided in Table 2.1. For the small NaPCP users the soils analysed for dioxin were composites from sampling to a depth of 0.2 m whereas the PCP depth concentration profiles for the three sites investigated indicated that PCP concentrations extended to approximately 0.5 m below the soil surface. It is well documented that the dioxins and furans, and particularly the more highly chlorinated heptachloro and octachloro congeners, are significantly less mobile in soil than PCP. On this basis, the dioxins were unlikely to have migrated deeper than the upper few centimetres of the soil profile, and it was assumed that they would be completely contained within the 0.2 m sampling depth.

Table 2.1 Parameters used to estimate the total dioxin reservoir in soil at timber treatment sites

* Based on site-specific data ** Based on assumptions.

For the very large NaPCP user, the depth of contamination was estimated from a plot of concentrations as a function of depth for each of the affected areas listed in Table 2.1. The surface concentrations given in the table were then used to calculate the average soil concentrations up to the depth at which the dioxins were estimated to have declined to zero.

From the soil concentrations, and the affected area and depth, the total dioxin burden was estimated to be 0.15 g I-TEQ per site for small NaPCP users and 19.2 g I-TEQ for the very large NaPCP user (Table 2.2). Unfortunately, no site information was available for the moderate NaPCP users or the large NaPCP users with respect to either the concentration of dioxin residues present at the site or the likely areas affected by dioxin contamination. Soil burdens for these sites could not therefore be calculated directly. Using the soil burdens for the small NaPCP users and the very large NaPCP user previously described, however, the dioxin burdens for the moderate NaPCP users and large NaPCP users were estimated from a linear regression. On this basis, the total dioxin burden was estimated to be 0.77 g I-TEQ per site for moderate NaPCP users and 2.7 g I-TEQ per site for large NaPCP users (Table 2.2).

Previous national reservoir estimate for NaPCP-derived dioxin

For 270 small users, 11 moderate users, four large users and one very large user, the total dioxin reservoir in soils at timber treatment sites from antisapstain use of NaPCP was estimated to be approximately 80 g I-TEQ, as summarised in Table 2.2.

Table 2.2 Dioxin reservoir in soil from antisapstain treatment (NaPCP)

* Dioxin burden = (surface concentration/2) x affected area x depth x soil density. The surface concentration is divided by 2 to provide for a linear decrease to zero over the depth of the contamination. This is applied for the Waipa site only, not for Hanmer. The concentrations, affected area and depth are given in Table 2.1. A mean soil bulk density of 1.2 tonne m-3 was assumed.

** Dioxin burdens for the moderate NaPCP users and the large NaPCP users were obtained from a linear regression of the data for small and very large PCP users.

Preservative use of PCP in diesel

PCP as a preservative in diesel oil was only used at four sites in New Zealand: the Waipa Mill (where antisapstain treatment also occurred), Hanmer Springs, Christchurch, and a fourth site in Waikoau (Hastings District) which was a comparatively small user. The usage of PCP at the three major sites is estimated to be approximately 2,000 tonnes at the Waipa Mill, 420 tonnes at Hanmer and 275 tonnes at Christchurch (although no dioxin data are available for the Christchurch site).

Concentration data for dioxins at the Waipa site have been reported, and are summarised in Table 2.1. The Hanmer Springs site estimate in Table 2.1 is based on assumptions extrapolated from the Waipa site. The mean dioxin concentration assumption cannot be verified since contaminated soil in the vicinity of the treatment plant was removed and disposed to an unknown location. The estimated areas potentially affected by dioxin residues at both of these sites are also detailed in Table 2.1.

As previously noted for the Waipa site, the depth of contamination was estimated from a plot of dioxin concentrations as a function of depth for each of the affected areas listed in Table 2.1. The surface concentrations given in Table 2.1 were then used to calculate the average soil concentrations up to the depth at which the dioxins were estimated to have declined to zero. On this basis, the depth of contamination at Hanmer was assumed to be the same as that at the Waipa Mill.

Previous national reservoir estimate for dioxins derived from PCP in oil

The estimated burden of preservative-derived dioxins at the Waipa Mill is estimated to be 184 g I-TEQ and at Hanmer 45 g I-TEQ, as summarised in Table 2.3. The total dioxin reservoir in soils at timber treatment sites from preservative use of PCP in oil is estimated to be approximately 230 g I-TEQ.

Table 2.3 Estimated dioxin reservoir in soil from preservative treatment

* Estimated from Waipa Mill data

2.3 Level of confidence in site contamination characterisation and reservoir estimates of the previous study

The level of confidence of the dioxin reservoir estimates for the very large user (PCP in oil and NaPCP) and the large user (PCP in oil) is relatively high, because both sites have been subjected to extensive investigation. The level of confidence of the dioxin reservoir estimate for small NaPCP users is very low because there is only a small amount of data, and this is for composites that were sampled to a depth of 0.2 m and from a limited number of locations. The level of confidence in the dioxin reservoir estimate for medium and large NaPCP users is also very low because mean dioxin burdens have been inferred.

The level of confidence in the dioxin reservoir estimate for PCP in oil sites is medium. There is some uncertainty about the quantities of PCP that were actually used at the sites, but there is good characterisation of dioxin contamination at the two major sites. There is also some uncertainty regarding the extrapolation of the contamination data to a dioxin reservoir estimate for the Hanmer Springs site. Although no dioxin investigations have been undertaken at the third significant user site, PCP investigations have shown only low levels of contamination and it is inferred from this that dioxin contamination will be similarly low. (As the results from the current study suggest, however, this may not necessarily be true.)

The level of confidence in the dioxin reservoir estimate for NaPCP use is low. The quantity of NaPCP use of many of these sites is uncertain and the soil dioxin concentrations and area of contamination are based on only a limited amount of site-specific data, which have then been extrapolated over the whole sector.

A desire to achieve better estimates of the overall dioxin reservoirs at sawmill sites was the major impetus for the current study.