7.0 Groundwater

7.1 Introduction

• the groundwater status;

• whether the groundwater is in compliance with the target criteria set out in the resource consent; and

• the implications of residual groundwater contamination on the receiving environment.

Part of this assessment includes determining whether the groundwater data are representative of actual conditions i.e. assessing the data quality. In addition, the hydrogeological conceptual models put forward in various earlier reports are commented on and potential information gaps identified. Finally, the environmental risks associated with residual groundwater contamination are examined.

7.2 Geology and Hydrogeology

Summaries of the geology and hydrogeology are presented in T&T (2003) and more recently CH2M HILL (2007). The following draws from these documents.3

7.2.1 Geology

The original geology at the site is presented as interbedded estuarine deposits (Rabbit Island Gravels and Tahunanui Sand) over the clay-bound Moutere Gravel. The top of the Moutere Gravel dips from west to east across the peninsula at the site location. Cross-sections in the AEE indicate the top of this unit is at less than 1m depth adjacent to the creek in the west, and approximately 6.5 m in the east (T&T, 2003b).

Some modification of the geology of the site had occurred prior to the onset of remediation. This included:

• the emplacement of a landfill to the west of FCC West. This included the construction of a clay bund along the north-western boundary (adjacent to the creek) and along the southern edge of the landfill (adjacent to the west beach). The exact position, depth and integrity of the clay bund is not known;

• the construction by TDC in the mid 1990s of a soil/bentonite cut off wall along the southern boundary of the landfill area;

• extensive land reclamation area along the seaward side of the East FCC ;

• land reclamation in the 1950s or 1960s along the seaward side of FCC East, adjacent to Mapua Channel, including construction of a clay bund/wall on the foreshore.

Since the commencement of remediation (October 2004), significant alteration of the surface geology across the site has occurred:

• surface soils have been excavated (typically to between 0.5 and 3 m) and replaced with fill material comprising treated soil, untreated residential and commercial quality soils, crushed concrete and “oversize” (>10 mm) material. Excavations extended below the groundwater table at many locations. The fill material has been compacted using different methods for different parts of the site;

• the base of much of the fill in FCC East and FCC Landfill consists of a layer of screened oversize material or crushed concrete.

• a system of interconnecting drains (so-called “French drains”) consisting of oversized material around the periphery of each backfilled subgrade in FCC East

• an earth bund consisting of Moutere Gravel has been put in place along the boundary between FCC Landfill and FCC West. The bund was installed to depths of between 2.5 and 4 m bgl and is founded in the Moutere Gravel formation (MWH, 2009f). This is at variance with the remediation as-built drawings, which show clay founded on marine sediment backfill. However, photographs in MWH (2009f) show subgrades SG19A, SG19B and SG19C having been excavated down to the Moutere Gravel formation;

• a number of service penetrations through the clay bund along the eastern foreshore were discovered during the remediation works and these were repaired. The removal of the surge chamber at this location resulted in a section of the clay bund being removed, with the excavation extending to a depth of over 5 m. However, the bottom 2.5 m of this excavation was backfilled with sand and gravel rather than clay (MWH, 2008b). Consequently, the eastern clay bund is discontinuous as a barrier to groundwater movement.

Layering of different fill types and the depth to the base of these fill layers is variable across the site. This has included placement of crushed concrete, screened oversize material and treated fines (which has been ground to clay-size particles during the MCD process) below the watertable in various locations. The oversize material and crushed concrete probably have a greater hydraulic conductivity than the material they replaced while the treated fines probably have a lower hydraulic conductivity than the native materials. Testing of the hydraulic conductivity of replacement materials has not been carried out.

7.2.2 Hydrogeology

The sediments overlying the Moutere Gravels form a shallow unconfined aquifer. The Moutere Gravel acts as a confining layer restricting vertical groundwater movement between the shallow aquifer and deeper aquifers. The Waimea Inlet and Mapua Channel form natural boundaries to the aquifer and represent aquifer discharge zones. Recharge to the aquifer is expected to be predominantly from the land mass to the north but also vertical infiltration and leakage from the underlying Moutere Gravels.

Discharge of groundwater occurs to the drain (the creek) excavated along the northwest boundary of the landfill (and from there to the Waimea Inlet), direct to the Waimea Inlet at the beach south of the Landfill and to the Mapua Channel at the FCC East beach. A seepage line on the beach south of the Landfill 2 – 3 m below high tide level was reported by T&T (2003). The reclamation to the east, the bentonite cut off on the boundary of the landfill and the newly installed bund between FCC West and the Landfill are likely to affect groundwater discharges to the Mapua Channel and Waimea Inlet.

Based on groundwater level measurements, Woodward-Clyde (1996) produced a groundwater contour plan which showed discharge at the northern end of the East beach and towards the south-eastern tip of the Landfill. Groundwater was interpreted to be flowing under the cut off wall at this point. A groundwater divide was inferred to run diagonally north-west to south east across the centre of FCC West. Hydrogeological parameters (permeability and porosity) for the aquifer pre-remediation have been estimated as (T&T, 2003b):

• hydraulic conductivity – 1 to 20 m/d; and

• porosity – 0.2 to 0.3.

Calculated linear (or seepage) groundwater velocities were reported in CH2M HILL (2007) to be in the range of 0.04 – 0.55 m/day for FCC East and 0.018 – 0.23 m/day for FCC Landfill. The substantial excavation and backfilling may have altered the bulk hydraulic properties of the aquifer and these values may not still be valid.

Pre-remediation groundwater levels of 1.5 – 2 m below ground level across the site are reported in the AEE (T&T, 2003b). A somewhat greater range of water depths (1 – 2.7) was reported towards the end of the remediation by CH2M HILL (2007) although greater depths would be expected as the ground surface is higher post-remediation. PDP (2009) presents plots of water levels over five years showing a seasonal range of 1 m in monitoring wells BH1 and BH5 near the eastern and western extremities of the site (Figure 1).

CH2M HILL (2007) comments on tidal influences on groundwater levels (based on observations in BH1 and BH2) and indicate response to a 3.7 m tidal range at the time of 0.26 m at 10 m from the shoreline and 0.035 m at 50 m from the shoreline.

Two slightly different interpretations of groundwater flow directions have been derived from groundwater level measurements in on- and off-site wells post-remediation. CH2M HILL (2007) indicates a groundwater divide runs slightly to the west of Tahi Street, running down the centre line of the peninsula, with groundwater flow both westward and eastward of the divide (Figure 6). Recharge to the site and the peninsular south of the site is inferred to be predominantly from infiltration. CH2M HILL (2007) shows westward and southward groundwater discharge to the creek and FCC West beach and eastward discharge to the FCC East beach.

The CH2M HILL interpretation has been criticised on the basis that groundwater measurements were taken both before and after a heavy rainfall event so that the measurements were not a “snapshot” in time, but represent a range of groundwater conditions over a period when the groundwater level rose due the heavy rain. PDP (2007) has reinterpreted the data to arrive at a contour diagram that is more consistent with pre-remediation flow direction interpretations of Woodward-Clyde (1996) and T&T (2004). Both Woodward-Clyde and T&T (apparently a reinterpretation of the 1996 Woodward-Clyde data) used many more water level monitoring points to construct their contours than are available now.

The PDP (2007) reinterpretation of the CH2M HILL data (Figure 7) shows a southerly component to groundwater flow crossing the southern boundary of the site in addition to the eastward and westward components. This difference in interpretation is important because if no southerly component exists, private bores in various properties along Tahi Street should not be at great risk from contaminants migrating from the site. If a southerly component exists, the off-site wells could be at risk.

CH2M HILL (2007) has estimated groundwater flux through the site to the west and east to be 1.6 to 42 m3/d and 10.1 to 154 m3/d, respectively, based on the groundwater gradients derived from their 2007 groundwater investigation and permeability and porosity data provided by Woodward-Clyde. However, the groundwater flux calculations are in error by factors of between 3 and 5 as these calculations should have used the Darcy velocity rather than the seepage velocity (the velocity from the Darcy equation divided by porosity). CH2M HILL’s values are compared in Table 6 with correctly calculated values and further groundwater flux estimates using hydraulic gradients from PDP (2007) for water level measurements taken in November 2007.

Two different estimates for the hydraulic gradient for FCC East were used in the revised calculation; one using the measurements from BH2 and BH1A (as CH2M HILL had done) and a second calculation using the contours further north, to better represent the average hydraulic gradient in the eastern part of FCC East.

The corrected and new calculations show a range of groundwater flux estimates and demonstrate the considerable uncertainty with the estimates of both the hydraulic gradient and hydraulic conductivity.

The remediation has changed the shallow geology and potentially the groundwater behaviour, but the effects on the groundwater are difficult to predict and cannot be assessed from the available information. Possible changes include:

• creation of preferential flow paths and/or local flattening of the hydraulic gradient where crushed concrete or oversized material has been placed below the watertable. Similar effects could be expected from the drains around each subgrade in FCC East;

• creation of lower permeability zones where the treated fines have been placed below the watertable;

• changes to the infiltration characteristics of the ground as a result of excavation and backfilling over most of the site and changing the surface cover from sparse vegetation to grass; and

• changes of flow patterns as a result of installing a clay barrier between FCC West and the landfill area and changes to the clay barrier on eastern foreshore.

The effect of these uncertainties is discussed in more detail in Section 1.18.

A better understanding of flow directions and hydraulic gradient requires the installation of more monitoring wells on the site, particularly in the central part of FCC East and within FCC west, including wells on the upgradient boundary. It would be helpful to undertake slug tests in existing and new wells to obtain hydraulic conductivity estimates of the material now in place.

7.3 Groundwater Contamination Sources

Prior to the remediation, site investigations identified the primary contaminants of concern to be organochlorine pesticides DDX and ADL associated with the site’s original use. These remain as contaminants of concern for groundwater. Groundwater criteria were derived to protect the marine environment and became a condition of consent during the remediation. During the remediation several other contaminants became of concern as a result of the remediation, specifically:

• The nutrients ammonium, nitrate and phosphorus associated with the use of diammonium phosphate (DAP) and urea as reagents in the MCD soil remediation process. No groundwater criteria were set in the consent for these determinands as it was not realised that large quantities of DAP would be used during the remediation, although the use of urea was known.

• Copper associated with copper sulphate, and iron in an unknown form, also used as reagents in the MCD soil remediation process. Although copper sulphate was not listed as a reagent in the AEE, an SAC was set within the consents for copper as part of a suite of metals because some heavy metal contamination existed on the site. There was no SAC for iron.

As discussed previously, the remediation generated a number of different material types that were subsequently used as backfill on different parts of the site. Materials with higher contaminant concentrations, conforming to the commercial SAC, were permitted to be buried below 0.5 m depth on FCC Landfill and FCC East. In addition, residual contamination exists at concentrations conforming to commercial soil quality in the undisturbed soil below the remediated areas within FCC East and Landfill. These materials are expected to be the principal ongoing sources of contamination for groundwater on the site. Generation of leachate will occur through vertical percolation of rainfall recharge and, where the contaminated soil is below the water table (either permanently or intermittently) by horizontal migration of groundwater through the fill.

The commercial fill material consists of three basic types – commercial soil, treated fines, and oversize. Commercial soils are those which have not undergone treatment as the concentrations of contaminants were found not to exceed the commercial SACs at the time of excavation, whether within the excavated material or the base of the excavation.

Treated fines are those that have undergone MCD treatment to reduce concentrations below the commercial SAC. The oversize material was the greater than 10 mm fraction that was removed from soil to be treated before it was processed. Typically, the treated fines were mixed with commercial soil and/or oversize before backfilling (MWH, 2009e). However, in some cases unmixed treated fines were placed directly in the excavation. Where an excavation was below the groundwater table and there was standing water at the time of backfill, coarse material such as crushed concrete and/or oversize was end-tipped to form a platform.

Contaminants at excessive concentrations in the soil could generate sufficient leachate that groundwater concentrations would exceed those thought appropriate at the point of discharge of the groundwater to the marine environment. The SACs were derived so that groundwater concentrations would not be excessive (Egis, 2001). The Site Auditor (GHD, 2006b) stated that the broad intention of the remediation was to place commercial quality soil above the water table. It was recognised that there was uncertainty in the derivation of the SACs and that placing commercial soil above the water table was one way of reducing this (Peter Nadebaum, GHD, pers. comm.) In a similar manner, it was assumed that infiltration would be reduced by vegetation and paving. It is obvious that commercial quality material can be expected to have greater leaching potential if placed below the watertable as it is subject to constant leaching from the groundwater flow rather than intermittent leaching from rainfall infiltration.

However, the intention to place commercial quality soil above the watertable was never formalised in the consents, with nothing in the consent preventing material at concentrations meeting the commercial SACs being placed below the watertable. In fact the original Thiess (2004) RAP had an explicit expectation that treated material would be placed below the watertable in stating:

Where material is to be placed below the water table (i.e. into standing water)… [it] will have a similar permeability to pre-existing material, achieved through the combination of treated material with screened oversize.

This was changed in the MfE RAP, with the record of changes to the RAP (MfE, 2007) showing that Work Plan 9 was changed in December 2006 to explicitly require all treated material and commercial grade material, including any oversize, to be placed above the watertable. It would appear that this requirement either came too late in the remediation or was not implemented properly.

The locations of the various materials and the depths to which they were placed are shown in the MWH as-built drawings (MWH, 2008b). The PDP (2007) groundwater issues report compares the estimated watertable elevation with the depth of the various commercial backfill materials to assess whether they could be below the water table. The results show that in a number of areas the treated fines, treated fines mixed with other commercial quality materials, and commercial soil without treated fines are either intermittently or permanently below the water table. As noted in PDP (2007), this estimate was based on uncertain groundwater elevation data which would require a more comprehensive monitoring network to increase certainty.

As noted above, in addition to the historic contamination residues, the MCD treatment process introduced potential contaminants. PDP (2007) reported that within the 20,969 tonnes of treated fines produced between October 2004 and July 2007, there was an estimated 1,970 tonnes of additives comprising sand (57.8%), diammonium phosphate (37%), copper/iron (3.4%) and urea (1.8%). Testing of the treated fines has confirmed the presence of considerably elevated concentrations of nitrogenous compounds and copper.

The average copper concentration in the treated fines is approximately 1,400 mg/kg, significantly above background concentrations, although below the SAC for commercial soil of 5000 mg/kg. Similarly, the average total nitrogen concentration in the backfilled treated fines is about 5,000 mg/kg. Average DDX and ADL concentrations in the backfill materials are considerably lower, and also below their respective SACs of 200 and 60 mg/kg, as discussed earlier, but still constitute significant masses which provide ongoing sources of groundwater contamination, depending on their location and leachability.

Multiplying average concentrations with the mass of treated material results in mass estimates for the various contaminants remaining in backfilled treated fines and commercial soil (Table 7). There is a discrepancy between the amount of copper calculated from the measured concentrations in the treated fines and the amount of copper and iron that EDL reported adding to the process (PDP, 2007). Using the PDP (2007) figures, copper (as copper sulphate) was at most 3.4% of 1,970 tonnes of additives. This is the equivalent of about 17 tonnes of elemental copper, approximately half the amount calculated from the average concentration. Given the number of samples (159) used to obtain the average concentrations in the treated fines, it would appear that the amount of copper reagents added has been under-reported.

It can be seen that a considerable mass of nitrogen and copper compounds remains, with lesser amounts of DDX and ADL.

Synthetic precipitation leaching procedure (SPLP) tests¹⁸ provide an indication of the leaching potential of contaminants in soil. Such tests performed on soil samples taken from buried backfill indicate leachate concentrations of DDX, ADL and ammonia orders of magnitude above the PETC and ANZECC marine trigger levels (PDP, 2007), demonstrating the potential for leaching. The data for copper show leachate concentrations of the same order as the PETC.

The testing indicated that contaminant concentrations in leachate are not significantly affected by the type of backfill (treated fines or commercial) but are affected by soil contaminant concentrations. The SPLP results also shows a relative decrease in leachability from nitrogen compounds to ADL compounds and copper, and finally DDX (PDP, 2007).

The SPLP leachate concentrations were compared with groundwater concentrations in PDP (2007) to obtain a sense of how the potential leachability as measured by the SPLP tests is converted to actual groundwater contamination measured in monitoring wells. There is considerable uncertainty in this comparison because of such things as differences between wells in their spatial relationship between the treated material and the wells, whether the treated material is submerged or not near a particular well, the flow direction relative to the treated material and the well, and the infiltration characteristics of the “catchment” for the particular well.

However, PDP (2007) estimated that groundwater concentrations of OCPs were on the order of 10 to 50 times lower than the SPLP leachate concentrations, presumably reflecting the absorption of OCPs onto the aquifer materials. Non polar hydrophobic organic compounds such as DDT and dieldrin are known to adsorb to sediments, having high partition coefficient values¹⁹. In contrast, ammonia and nitrate are highly soluble and not readily absorbed onto sediments. Ammonia-N concentrations have been measured in groundwater at similar concentrations to the SPLP leachate values and nitrate-N has been measured at significantly higher values than the SPLP leachate test concentrations. PDP (2007) suggested the latter could be a result of some ammonia-N converting to nitrate-N.

Comparison of the leaching tests with groundwater concentrations for copper indicated attenuation of over 100 times. This is not unexpected given the tendency of copper to be absorbed onto the surface of fine soils.

7.4 Consent Monitoring Requirements

Requirements relating to groundwater are covered in conditions 23 to 32 of Consent RM030524. As noted previously, the consents for the project expired in November 2007.

The consent required the monitoring of six on-site bores (BH1, BH2, BH3, BH4, BH5, and BH9) and a minimum of four off-site bores, including 13 Tahi Street, 17 Tahi Street, 26 Tahi Street and 36 Tahi Street. BH1 is on the eastern boundary of FCC East, adjacent to Mapua Channel; BH2 is on the southern boundary of FCC East, adjacent to 13 Tahi Street; BH3 is at the south-western extremity of FCC Landfill; BH5 and BH4 are on the southern and north-western boundaries of the landfill respectively; and BH9 is on the southern boundary of FCC West, adjacent to 18 Tahi Street. The wells are shown on Figure 1.

A number of additional monitoring wells were installed during the CH2M HILL (2007) investigation in May 2007. Several of these were intended as replacement wells for the original monitoring network and were named BH1A, BH2A, BH3A, BH4A, BH5A, and BH9A (see Figure 1). These are generally within a few metres of the original wells except for BH1A, which is around 20 m from BH1. It appears that BH2 was the only monitoring well listed in the consent to be destroyed during the remediation works.

Monitoring of groundwater in the likely upgradient direction to obtain background concentrations was not specified in the consent.

The minimum suite of parameters to be monitored was specified in the consent as an OCP suite, a metals suite, acid herbicides, electrical conductivity (EC), pH, alkalinity, and static groundwater levels.

The consent specified monitoring of on-site wells prior to and during the remediation, with a monthly frequency for the duration of the works. Monitoring of off-site wells was specified prior to and following commencement of remediation, with a frequency of every three months during the remediation.

Based on advice from the Site Auditor, and with the agreement of TDC, the analytical suite was altered a number of times during the remediation works. In January 2005, ONP and OPP were added to the suite of analytes for all wells, and volatile organic compounds (VOC) were added for the on-site wells. In June 2005, the analytical protocol was changed to become:

• monthly – OCP, EC, pH, total alkalinity and static groundwater levels;

• quarterly – the monthly parameters plus: total kjeldahl nitrogen (TKN), nitrate, total phosphorus, copper and carbaryl (an OPP); and

• annually – the monthly and quarterly parameters plus: VOC, acid herbicides, OPP, ONP and a metals suite.

In October 2005, ammoniacal nitrogen, nitrate and nitrite were introduced on a monthly basis for all on-site wells and for every three-monthly monitoring of the residential wells. Condition 28 of the resource consent sets out Provisional Environmental Threshold Concentrations (PETC) for groundwater as follows:

The PETC were derived as trigger values to be applied during the remediation works. If the values were exceeded during the remediation, the consent required additional investigation to determine the source of contamination and implementation of corrective measures.

It is understood that the PETC presented in the table above are intended to be those derived by T&T and presented in Table 8.2 of their Groundwater Assessment Report of May 2003 (T&T, 2003b). The PETC presented in T&T’s report were quoted as being based on l00 times the relevant ANZECC 99% Freshwater Level of Protection (LOP) value (ANZECC/ARMCANZ, 2000). The factor of 100 allowed for the 100-fold dilution of groundwater which was estimated to occur in the Mapua Channel.

There are several points which should be noted:

• The PETC presented in both the T&T report and consent are not based on the ANZECC 99% freshwater LOPs, but for the most part are based on the marine low reliability trigger values (ANZECC/ARMCANZ, 2000). Exceptions are as follows: DDD (which has never had an ANZECC trigger level); dieldrin (combined marine/freshwater low reliability trigger value); mercury (which is based on the 95% marine LOP divided by 1,000 to allow for bioaccumulation); copper (95% marine LOP); selenium and zinc (which are the only ones based on the 99% freshwater LOP); and total chromium (for which the hexavalent chromium (Cr VI) marine 95% LOP has been used). The basing of the PETC on marine values is reasonable given the intent is to protect the marine environment and the reference to freshwater guidelines in the T&T report appears to be an error;

• No allowance is made for bioaccumulation of DDX, aldrin, dieldrin, chlordane, heptachlor, and selenium;

• Whilst it is understood that allowance has been made for 100-fold dilution in the derivation of the PETC, the value shown in the consent for DDT actually represents 1000-fold dilution of the quoted ANZECC trigger.

• Whilst a figure of 100-fold dilution was used to derive the PETC (except for DDT) the level of dilution for groundwater entering the Waimea inlet was estimated by T&T (2003b) to be only five-fold. However, CH2M HILL (2007) later estimated dilution in the Mapua Channel and Waimea Inlet to be of the order 80,000 – 1.5 million and 1000 – 25,000, respectively.

• The table of PETC in the consent states that the values relate to the assumption of a cut-off wall along the eastern boundary of FCC East. Apart from DDT, this is the same as the table for proposed Condition 55 in the AEE (T&T, 2003a). However, the same values presented in Table 8.2 of the T&T groundwater report state that they relate to the assumption of no cut-off wall along the eastern boundary of FCC east. If a cut-off wall is assumed, the values should have been ten times higher in the consent (except for DDT which was ten times higher), if Table 8.2 of the T&T groundwater report is accepted.

These apparent discrepancies in the derivation of the PETCs have no particular bearing on the application of the consents as the values in the consent prevail. However, in considering the post-remediation effects it is appropriate to adopt a consistent dilution factor for the various contaminants.

7.5 Additional Monitoring by TDC

When the consent expired in November 2007, TDC continued monitoring a reduced number of wells for selected analytes on a three-monthly basis. The wells monitored by TDC included: BH1/BH1A, BH2A, BH5/BH5A, BH9/BH9A and the residential bore at 13 Tahi Street. Where a well pair exists, in some of the earlier TDC monitoring rounds both wells were sampled, but more recently only the ‘A’ wells have been sampled.

The parameters included in the TDC monitoring are: nitrate nitrogen, ammoniacal nitrogen, total nitrogen, dissolved reactive phosphorus, copper, iron, DDX and ADL.

In November 2008, the monitoring by TDC also included (see Figure 1):

• BHD, in the northeast corner of the site, adjacent to Mapua Channel;

• BHH on the southern boundary of FCC East, adjacent to 13 Tahi Street;

• BHG on the southern boundary of FCC West, adjacent to 18 Tahi Street;

• BH3A and BH4A, along the north-west boundary of FCC Landfill;

• Old BH1 on Tahi Street near the centre of the northern site boundary;

• BHL on Tahi Street approximately 50 m south of the site boundary; and

• eight additional residential bores in Tahi Street properties to the south of the site (17A, 21, 23, 26, 27, 29, 36 and 39 Tahi Street).

7.6 Post remediation groundwater criteria

Groundwater monitoring results post-remediation have been compared with both aquatic ecosystem guidelines (ANZECC, 2000) and the Drinking-water Standards for New Zealand (MoH, 2005). CH2M HILL (2007) used human health and aquatic guidelines for ammonia, nitrate and nitrite but used the consent PETCs for OCPs. PDP (2007), in reporting the monitoring by TDC have variously used aquatic guidelines and drinking-water MAVs (maximum allowable value) and GVs (aesthetic guideline values). The various criteria are summarised in Table 9.

The groundwater presents potential risks through:

• human consumption of groundwater;

• use of groundwater for irrigation;

• industrial use of groundwater;

• discharge of groundwater to an area of human contact recreation in the estuary and Mapua Channel; and

• discharge of groundwater to the marine ecosystem

It is appropriate to compare with human health guidelines where people might come into contact with the groundwater long-term, whether in drinking the water or in contact recreation (where the drinking-water guideline is used as a proxy). It is understood (CH2M HILL, 2007) that none of the private bores in Tahi Street are used for drinking purposes, being used primarily for irrigation. However, there is nothing stopping owners using the water for drinking purposes. In addition, there is a potential for further bores to be constructed and water taken. Tasman District Council’s Tasman Resource Management Plan has the following provisions with respect to installing a bore and taking small quantities of water:

• a bore permit is required for a drilled well, but not for a ‘dug’ well (including by excavator) to a depth of 8 m;

• TDC can only reject a bore permit application on spacing grounds. The minimum spacing for the shallow aquifer is 50 m; and

• taking up to 5,000 L per day for domestic purposes is a permitted activity;

A plan change would be required to control potential abstractions on or adjacent to the site, for example to make such abstractions a discretionary activity for a specific area.

While there is no particular reason why a resident would go to the expense of installing a bore, given the existence of a reticulated supply, there is nothing theoretically stopping such use. Consequently, a potential risk exists if the groundwater is contaminated above drinking-water criteria. Given this, it seems reasonable to compare the groundwater sampling results with human health criteria. This is a more sensitive use than (garden) irrigation use.

Protection of the marine environment has been a primary objective of the remediation. Comparison with the PETCs is intended as a measure of the threat to the marine environment. The PETCs are intended to be protective of the marine environment after mixing, with, as noted above, a 100-fold dilution generally assumed in their derivation (1000-fold for DDT). Implicit in this is the acceptance that there could be some effects on the marine ecosystem at the point of discharge, where dilution would be less. Given this, it seems reasonable to compare the nutrient concentrations, at least with respect to their ability to be toxicants, with guidelines that recognise dilution. This has not been the case to this point in reports on the groundwater to date.

This then raises the question as to what dilution is appropriate. CH2M HILL (2007) calculated dilution of 80,000 to 1.5 million in the Mapua Channel and 1,000 to 25,000 in the Waimea Inlet. These are several orders of magnitude higher than the dilution factors of 100 and five respectively for the channel and inlet calculated by T&T in the AEE (although the greater dilution factor was used to derive the PETCs). The different estimates reflect the different methodologies, with T&T basing its estimates on flow rates past the discharge zones and a 20 m mixing zone, and CH2M HILL basing its estimate on complete mixing within the volume of water emptied during each tidal cycle.

The T&T estimates are very conservative and the CH2M HILL estimates are probably extreme. A more reasonable dilution estimate is perhaps in the range of one to several thousand, which should still be conservative for the channel but less so for the inlet. Inevitably, using larger dilution factors means there is a greater risk at the point of discharge at low tide (and within the pore water within the marine sediment). At least with respect to nutrients, there is a demonstrated excessive level of nutrients at the point of groundwater discharge as evidenced by enhanced algal growth.

The resource consent process did not appear to address the trade-off between local and general effects, except implicitly by accepting conservative dilution factors. Given the on-going groundwater contamination this needs to be considered more explicitly, but requires more information. Clearly there is algal growth from excessive nutrients where groundwater discharges on the east and west foreshores. However, there is no information on the potential effects on the beach/mudflat ecosystems of undiluted or partially diluted groundwater discharges when the tide is out. Meanwhile, in reviewing the results of monitoring to date, comparisons have been made here with aquatic criteria assuming 1,000 – 5,000-fold dilution as reasonable intermediate values between the T&T (2003a) and CH2M HILL (2007) extremes.

Before establishing such guidelines and making the comparisons, it should be noted that CH2M HILL (2007) quotes guidelines for nitrate and nitrite (as N03 and N02, respectively) in its report yet the results were reported by the laboratory as nitrate-N and nitrite-N. CH2M-Hill did not convert the guidelines to be in terms of nitrogen alone and therefore their comparisons were incorrect in an un-conservative sense. The effect of this is minor if the results are compared with conservative undiluted marine ecosystem guidelines, as the results are generally so high as to well exceed the guidelines, but the error may be significant if dilution is taken into account.

CH2M-HILL also used a low reliability marine nitrate guideline (as a toxicant not nutrient) from ANZECC (2000) which is based on the 95% protection level freshwater guideline. ANZECC (2000) states:

Although a marine low reliability figure of 13 000 µ g/L (13 mg/L) could be calc ulated using an AF of 200 (limited data but a lesser factor due to essentiality), it is preferable to adopt the freshwater figure of 700 µ g/L for nitrate toxicity as NO 3 (nitrate) as a marine low reliability trigger value.

However, in 2002, MfE issued a memorandum from the National Institute of Water and Atmospheric Research (NIWA, 2002) which questioned the validity of the 700 µg/L – this value was apparently in error – and derived a corrected value of 31,900 µg/L using the data presented in ANZECC (2000). This revised value is the equivalent of 7200 µg/L nitrate-N. If this value was to be used as a low reliability marine guideline, a somewhat different view of the groundwater monitoring results would result. It is not known whether such a value is appropriate for the marine environment. Further investigation is required.

PDP (2007), in using the human health value for ammonia, incorrectly compares this guideline with ammoniacal-N results. The error is small given the small difference in formula weights. However, the value of 0.3 mg/L used by PDP is to guard against the formation of chloroamines in a chlorinated water supply. This has limited relevance in the current situation and does not appear in the 2008 update of the New Zealand Drinking-water Standard (MoH, 2008). Perhaps a more relevant value is the aesthetic guideline of 1.5 mg/L measured as the ammonium ion. This is the equivalent of 1.2 mg/L measured as ammoniacal-N.

Table 9 provides a comparison of various human health and aquatic guideline values, the latter using 100-fold, 1000-fold and 5000-fold dilutions. The 100-fold dilutions are similar to the PETCs.

7.7 Groundwater Monitoring Data Quality

7.7.1 General Scope and Methods

The groundwater monitoring wells can be divided into three categories:

• private bores intended for water abstraction;

• monitoring wells installed prior to the remediation commencing; and

• monitoring wells installed by CH2M HILL, some of which replaced (or duplicated) existing monitoring wells

Little is known about the private bore construction and their suitability as monitoring wells. Monitoring wells BH1 – BH5 were installed by Groundwater and Environmental Services Limited during the initial characterisation investigation (GES, 2002) and appear to have been constructed to accepted industry standards. The replacement ‘A’ wells installed by CH2M HILL also appear to have been constructed appropriately.

In both cases, the screens fully penetrate the aquifer and therefore collect a sample that will be some sort of average of the groundwater conditions at the point. This may or may not be a good representation of the conditions at the point, depending on relative permeabilities of layers that the well may intersect and relative contaminant concentrations within those layers. Multiple wells screened at several levels or multilevel low-flow sampling techniques in the existing wells could be employed in an attempt to determine whether particular layers have higher or lower concentrations. However the wells are adequate for the current purposes.

In general, the sampling methods applied appear to be in line with accepted industry practice. There is a suggestion, however, that the initial sampling of the wells by CH2M Hill occurred too early after the wells were installed. It is good practice to leave the wells at least a week or two after installation before sampling, but in the case of the first monitoring of the new wells it was only a few days. The effect of this is not known, but of no great significance, as it is the trends over time that are of most important rather than the results of a single sampling.

One aspect of the protocol that was the subject of discussion during the remediation was whether or not to filter the samples analysed for OCPs. The original consent specified that such samples were to be filtered. However, the method adopted during the remediation works was to sample directly into laboratory prepared bottles with a peristaltic pump and for laboratory to centrifuge the samples to remove sediment (this is noted on the laboratory reports). The subsequent sampling by TDC also sampled directly into bottles and the laboratory is reported to have decanted the sample before using a 0.45 µm filter prior to analysis. The CH2M HILL (2007) investigation report records that OCP samples were allowed to settle in the field before decanting.

It would have been desirable to have adopted the same technique for all groundwater sampling throughout the project to give greater certainty that the sampling results can be compared.

There is no absolute rule on whether to filter samples for pesticide analysis. The obvious objective in groundwater sampling is to determine the concentrations within the groundwater, however, the sampling technique can change these concentrations. Turbid water is one factor, because contaminants can be absorbed onto the sediment suspended within the sample. The sediment is typically an artefact of the sampling, rather than the true state of the groundwater. It is common to filter samples for metals analysis if dissolved concentrations are desired. However, groundwater sampling procedures typically recommend collecting pesticide samples, for which very low concentrations are typically being measured, without field filtering. This is because of the possibility of the filtering equipment absorbing some of the contaminant, resulting in a falsely low result on analysis. However, if the sample is turbid and the sediment has the contaminant adsorbed onto it, the analytical result would be falsely high.

A partial solution is to let the samples settle and decant the clear supernatant for analysis. This may still leave very fine sediment (colloidal particles). A better alternative is to remove the sediment in the laboratory by centrifuging. It is preferable, however, that the samples are not turbid in the first place, preferably less than 10 – 15 NTU. This can be achieved by constructing and developing the monitoring wells so that little sediment is produced when sampling, and taking great care to minimise the production of sediment during sampling. This can be difficult if the wells are installed in fine soil, but there are techniques available such as employing pre-packed screens or geotextile filter socks.

CH2M HILL recorded turbidity when sampling in 2007 and there appears to be a relationship between higher turbidity and higher OCP concentrations (PDP, 2007). For example, BH5A is typically quite turbid and varies from one sampling occasion to another. However, from the data presented, it is difficult to quantify this effect or assess its potential significance. In any case, the data obtained without filtering are likely to be conservative when compared with actual dissolved concentrations in the aquifer. The practice of measuring the turbidity of samples has continued since 2007 with the ongoing TDC sampling. It is a good practice and should continue with future monitoring.

Notes of the TDC sampling from 2008 indicate that some wells were purged dry, resulting in turbid samples. Avoiding pumping wells dry (which will stir up sediment) and leaving the wells to settle for some hours after purging before sampling would assist in reducing sample turbidity for future monitoring. However, recent sampling suggests most samples (except BH5A) had acceptable turbidity.

7.7.2 Well Location and Monitoring Frequency

The number of on-site wells monitored during and after the remediation works is low for assessing post-remediation conditions. Sites of similar size typically have many more wells. Given the large investment in the remediation to date, the investment in post-remediation monitoring seems low. A more robust monitoring network would enable a better sense of the groundwater flow directions and seasonal variations of flow direction and water level to be gained and a better sense of the variation of contaminant concentrations across the site. Of importance is to confirm the presence of a groundwater divide running roughly north-south through FCC West, to confirm that it is unlikely that contaminated groundwater is flowing from FCC East to FCC West. Further wells could also determine whether there is a southerly component of flow towards the properties on Tahi Street, and how significant this southerly flow might be from FCC East. Better mass flux estimates could also be calculated so that discharge to the marine environment, particularly to the east, is better understood. The addition of the following wells would result in a more robust network:

• at least one upgradient well in the vicinity of the northwest corner of the site to obtain background;

• a well midway along the western boundary of FCC West;

• a well in the south-east corner of FCC West;

• a well in the former vicinity of the MCD plant;

• a well roughly at the midpoint of the south-western quadrant of FCC West;

• two wells spaced out along the western boundary of FCC East;

• two wells spaced out along a line running north-south midway across FCC-East;

• a well in the vicinity of the former surge chamber location where there is a gap in the clay bund;

• ideally, wells either side of the clay bund installed on the eastern boundary of the FCC Landfill site; and

• in addition to those wells currently monitored, routine monitoring of Old BH1, BHH, BHG and BHD, at least for water level.

Not all wells would necessarily be monitored for contaminants on every occasion, but at least a baseline should be established over an initial year of quarterly monitoring. Possibly a sub-set would be monitored after that, but that would depend on what was found in areas not currently monitored.

As discussed earlier, the hydrogeology on the site is not straightforward, particularly with the potential influence of the various clay bunds on groundwater flow patterns on the site (see Section 1.11). In fact, it is not entirely clear where some of the monitored wells are in relation to the clay bunds. According to information from TDC, the three landfill monitoring wells (BH3A, BH4A and BH5A) are located within clay bunds.

7.7.3 Analytes

The range of analytes tested for is reasonable for characterising the groundwater quality and assessing the key potential contaminants of concern.

7.7.4 Quality Assurance and Quality Control

In general, the QA/QC information for the groundwater monitoring is limited and it is difficult to assess the data quality. In each case the laboratory used was Hill Laboratories and, as discussed in Section 1.1.3.2, the laboratory’s internal QA/QC procedures are expected to be in line with industry practice. However, no internal laboratory QA/QC reports were provided for the project.

The groundwater monitoring during the remediation works was undertaken by ChemSearch, Dunedin. QA/QC sampling in the form of a field blank and duplicate was completed for each monitoring round. However, no analysis of the QA/QC results is provided in the reports.

The CH2M HILL provides an adequate QA/QC assessment for the groundwater sampling during the 2007 investigation. The report discussed the significance of duplicate, rinsate and field blank samples and concludes the QA/QC criteria were met for the investigation (CH2M HILL, 2007). We concur.

No QA/QC information was presented for the post-remediation sampling undertaken by TDC.

7.8 Review of Groundwater Monitoring Results

Monitoring of on-site wells during the remediation showed increasing trends in groundwater concentrations for a number of the key contaminants. In general, there has been a rise in groundwater concentrations in most wells following the commencement of remediation. In many cases the rise was more than an order of magnitude. While there has been a decreasing trend since remediation was completed, the groundwater has not returned to pre-remediation concentrations, suggesting the remediation is having on-going effects on the groundwater quality.

The detected concentrations of DDX and ADL exceeded the PETC in a number of locations, and concentrations of nitrate, ammoniacal-N and dissolved reactive phosphorus (DRP) rose to high concentrations. It is not clear whether any action was taken with respect to PETC exceedances.

The CH2M HILL report (2007) summarises the key results during the remediation and for the monitoring it carried out. A series of reports by PDP summarise the results of the post-remediation monitoring undertaken by TDC. Time series data are presented in these reports for compounds and wells where concentrations are consistently above laboratory detection limits.

In some cases, the post-remediation monitoring by TDC included both the original wells and the replacement ‘A’ wells installed by CH2M HILL. BH2 was destroyed during the remediation and hence only BH2A could be monitored. Both the original and replacement wells (BH1/BH1A, BH5/BH5A and BH9/BH9A) were monitored for the January 2008 and April 2008 monitoring events (i.e. the first two post remediation events). From that point on only the replacement wells have been. This is reasonable as the two datasets appear similar. For the purposes of this discussion, a reference to a monitoring well such as BH1 includes data from both BH1 and BH1A.

The various contaminants are discussed in more detail below. In general, detailed comment has only been made on data for the wells monitored regularly since the remediation was completed (BH1, BH2, BH5, BH9 and 13 Tahi Street). These wells represent the concentrations that would be going off site if the groundwater flow is crossing the boundary at the particular locations, i.e. they represent downgradient conditions, but they are not necessarily representative of conditions for the particular boundaries. This is because proximity of submerged or intermittently submerged treated fines could result in higher results than other locations on a particular boundary and more permeable material buried at the base of excavation could cause anomalous high or low results, depending on spatial relationships with treated soil. For example, treated fines, mixed treated fines or commercial material were placed below the watertable or within the zone intermittently submerged in the following locations close to monitoring wells:

• in all the landfill excavation subgrades, within which monitoring wells BH3A and BH5A were installed; and

• in subgrades SG7 and SG9, upgradient of monitoring well BH1A.

In addition, mixed treated fines were placed above the watertable in subgrades SG14 and SG16. Monitoring well BH2 is installed on the boundary of these two subgrades.

Thus these wells might be expected to have higher concentrations than wells installed where there is less treated material or where the treated material is above the watertable. Currently there are no wells in the centre of either FCC east or FCC West, so there is no sense of how representative the current less than ideally placed boundary wells are. It could be that the concentrations measured in, for example, BH1A, are an overestimate of typical conditions for the eastern side of FCC East. The few measurements in BHD suggest that that might be the case.

7.8.1 Pesticide Monitoring Results

DDX Compounds

There appears to have been an increase in DDX concentrations in groundwater associated with the remediation works, with general rises in concentrations in early to mid 2005 (often by an order of magnitude). Typically, this was followed by slight declines, although concentrations have not returned to pre-remediation levels. Looking at the individual compounds in BH1, BH2, BH5 and BH9, it is generally DDD and DDT which are at higher concentrations and showing greater rises compared with DDE.

CH2M HILL (2007) states that there are no clear trends for DDX concentrations in off-site wells (with the exception of 13 Tahi Street). The lack of regular monitoring of other off-site wells means that any seasonal fluctuations cannot be determined. However the monitoring of several off-site wells²⁰ in November 2008 by TDC found only trace concentrations of DDX at 26 Tahi Street, with all other wells non-detect. The most recent monitoring at 13 Tahi Street (the closest residential bore to the site) indicates DDX concentrations have been below laboratory detection limits in recent monitoring rounds, and hence below the equivalent PETC.

A single monitoring event is not sufficient to show there is no threat to all off-site wells from DDX compounds (and by extension other properties south of the site) but the recent consistent trend in 13 Tahi Street suggests a similar trend may also exist for the other off-site wells.

In reporting the DDX compounds for TDC in the various quarterly monitoring reports, PDP does not compare the individual compounds against the PETCs, rather it compares the sum of the DDX compounds against human health guidelines. However, all on-site boreholes for which data are presented have exceeded the PETC for the individual DDX compounds at some point during the remediation works. Since May 2008, only BH1 and BH5 have exceeded the criterion. Both these wells are in or close to mixed buried fines material.

In the three most recent monitoring rounds (August 2008, November 2008 and February 2009), the DDT concentrations in BH1 and BH5 were less than twice the PETC of 0.0004 mg/L. Although BH2 has reached around twice the PETC on one occasion, it is currently running at less than the laboratory detection limit (0.00001 mg/L) and looks likely to continue to do so. Apart from a single spike in June 2006, DDT concentrations in BH9 are consistently below the PETC and look likely to continue to be.

A similar pattern is shown for DDD concentrations. In the three most recent monitoring rounds, the DDD concentrations in BH1 and BH5 were up to maximum of about five times above the PETC of 0.0006 mg/L. The DDD concentrations in BH2 and BH9 have remained below the PETC and look likely to continue to do so.

DDE concentrations in BH5 and BH9 consistently exceeded the PETC of 0.00005 mg/L during most of the remediation works, typically by approximately eight times and two times respectively. During recent monitoring rounds, the DDE concentrations in these wells have been closer to or below the PETC. The DDE concentrations in BH1 have been close to or below the PETC throughout the monitoring period. An exceedance of DDE by around two times the PETC occurred in early data in BH2, along with a notable spike in July 2006. However, the DDE concentrations in BH2 appear to have stabilised below the laboratory detection limits (0.00001 mg/L) in recent monitoring rounds.

If the DDX data is considered in light of less conservative dilution factors (remembering the DDT PETC is based on 1000-fold dilution), the recent (and most other) monitoring would have complied in all cases with 1000-fold dilution for DDD and DDE and 5000-fold dilution for DDT.

Apart from the off-site wells, the November 2008 monitoring round undertaken by TDC also included the following additional on-site wells, all close to site boundaries: BHG, Old BH1, BHH and BHD. These had similar DDX concentrations to those on-site wells that are regularly monitored. DDX concentrations in Old BH1 (close to the northern boundary) were below laboratory detection limits.

The November 2008 monitoring round also measured DDX concentrations in BH3 and BH4 on the landfill boundary. DDX concentrations in BH3 were above the PETC and similar to BH5, which is not surprising given the two wells similar relationship to treated fines material. BH4 also showed elevated DDX concentrations, but less so than either BH3 or BH5. While still within or close to treated fines material, BH4 is on the edge of the landfill with groundwater flow estimated to be parallel to that edge, which may mean the groundwater passes through less treated fines material before reaching BH4.

In comparing the on-site wells with the drinking-water MAV of 0.001 mg/L for DDX in the New Zealand Drinking Water Standards (MoH, 2005), the only wells that have consistently exceeded are BH1 and BH5. During recent monitoring rounds the DDX concentrations in these wells have been between one and three times the MAV, confirming the water is unsuitable for drinking-water purposes. BH3 exceeded the MAV in November 2008 but BH4 did not. No off-site wells have exceeded the drinking water MAV for DDX in any of the monitoring rounds.

ADL Compounds

BH1, BH2, BH5, BH9 and 13 Tahi Street all showed increases in the sum of ADL concentrations starting in mid to late 2005, with BH5 in particular showing a series of elevated peaks until about March 2006. From March 2006 until July 2007, the ADL concentrations in the monitored wells appeared to be approximately static or decreasing, although concentrations have not returned to pre-remediation levels. From September 2007 until present, the ADL concentrations appear to have increased in BH2 and BH5, although the concentrations have oscillated and no consistent trend is apparent.

Looking at aldrin and dieldrin separately, the effects are generally more marked and concentrations higher for dieldrin than for aldrin. At 13 Tahi Street there are no clear trends.

The three most recent monitoring rounds indicate that aldrin concentrations in BH1 are well below the PETC and look likely to continue to be so. The aldrin concentration in BH2 peaked slightly above the PETC of 0.003 mg/L in November 2008, but was below the laboratory detection limit (0.000005 mg/L) in the monitoring rounds immediately prior to and after that round. In BH5, aldrin levels have occasionally come close to the PETC during the remediation works, but in more recent data have been half the PETC or less and look likely to continue to do so. Aldrin levels in BH9 have always been less than a third of the PETC, frequently much less, and look likely to continue to meet the PETC.

The most recent monitoring data indicate that the dieldrin concentrations in BH1 are well below the PETC of 0.001 mg/L and look likely to continue to be so. The dieldrin concentration in BH2 has recently been close to or slightly above the PETC on each occasion, a general increase on earlier data which was typically half this level or less. In BH5, the PETC for dieldrin has been frequently exceeded (generally fluctuating between half and 1.5 times the PETC although much more in the second half of 2005) and this looks likely to continue. Dieldrin levels in BH9 are generally less than a half of the PETC and frequently much less. The most recent data for BH9 are a third of the PETC or less, and this well looks likely to continue to meet the PETC. The November 2008 monitoring of BH3 and BH4 complied with the PETCs for both aldrin and dieldrin

The most recent monitoring data indicate that BH1 is meeting the PETC of 0.0007 mg/L for lindane, although the concentration in August 2008 was close to exceeding the criterion. The lindane concentrations in BH2 in the last three monitoring rounds have consistently exceeded the PETC, and have done so since about October 2006. Prior to that time, the results were always below the PETC in BH2. The peak concentration was recorded in August 2008 at 0.012 mg/L (17 times the PETC); while the next two results have been successively lower, with the most recent result at 0.0059 mg/L (8.4 times the PETC). The lindane concentrations in BH5 have recently fluctuated from below the PETC to approximately five times the criterion. This is consistent with previous results in that well. The lindane concentrations in BH9 have fluctuated from below the PETC up to about five times the criterion, and look likely to continue to do so.

The lindane concentrations in the 13 Tahi Street bore have been consistently below the PETC of 0.0007 mg/L, with a maximum concentration detected of 0.00014 mg/L in April 2006 (about one fifth of the PETC). All results are also below the drinking water MAV (0.002 mg/L).

The additional wells along the site boundary included in TDC’s November 2008 monitoring round found similar ADL concentrations to those on-site wells that are regularly monitored. A dieldrin concentration of 0.0075 mg/L in Old BH1 exceeded the PETC of 0.001 mg/L. Other ADL compounds were also present in this well, although at concentrations below the PETC (and therefore MAVs). The only off-site wells where ADL concentrations were detected were 26 Tahi Street where a trace dieldrin concentration was detected, and in BHL where a trace aldrin concentration was detected. Both concentrations were well below the corresponding PETC.

If criteria were adopted with greater dilution than assumed in the PETCs, then all the ADL compounds would have complied in recent monitoring except the lindane concentrations in BH2 in August and November 2008, which would not have complied with 1000 times dilution but would have complied with 5000 times dilution.

The drinking-water MAV for the sum of dieldrin and aldrin (0.00004 mg/L) has consistently been exceeded in on-site wells. In BH1 and BH2 on FCC East, the concentrations have typically been about five to twenty times the MAV. In BH5 (FCC Landfill), the aldrin plus dieldrin (A+D) concentrations in recent monitoring rounds have been between 30 and 40 times the MAV. In BH9 (FCC West), the concentrations have recently ranged between five and ten times the MAV.

The drinking water MAV for A+D has also been consistently exceeded in the residential bore at 13 Tahi Street. In recent monitoring rounds, the concentrations were about two to five times the criterion. The data for other off-site wells is limited. However, in the November 2008 monitoring round, all A+D concentrations were below the MAV at the nine other off-site monitoring locations, and only two locations recorded A+D concentrations above the laboratory detection limits.

Chlordane and heptachlor

PETCs were derived for chlordane and heptachlor. Chlordane and heptachlor have not been detected above the laboratory detection limit of 0.000005 mg/L in any of the wells in recent monitoring events. These compounds do not qualify as contaminants of concern and will not be considered further.

Non-PETC Pesticide Compounds

Concentrations of various ONP and OPP compounds have been detected above laboratory detection limits in a number of wells, both on and off-site. However, the data are limited as these analyses were only included annually, with the last round completed in January 2007. In the on-site wells detected OPP and ONP compounds included atrazine (up to 0.003 mg/L), carbaryl (up to 0.001 mg/L), diuron (up to 0.001 mg/L) and simazine (up to 0.003 mg/L). Trace concentrations of various OPP and ONP compounds were also detected in the residential bores at 13 and 26 Tahi Street, and to a lesser extent at 17A Tahi Street. All detected concentrations in the off-site wells were below the corresponding drinking water MAV (MoH, 2005).

In a similar manner, trace concentrations of various acid herbicides were detected in a number of on-site wells and a single off-site well (26 Tahi Street) during the January 2007 monitoring. However, all concentrations were well below the corresponding MAV.

These compounds will not be considered further.

7.8.2 Metals Monitoring Results

Copper

Copper concentrations in the monitored boreholes have been below the PETC of 0.13 mg/L in all wells, except for a single exceedance (0.16 mg/L) in BH5 in the January 2007 monitoring round. Until the most recent monitoring round, the laboratory detection limit used was 0.01 mg/L. Concentrations in all wells have been less than 0.02 mg/L in the three most recent monitoring rounds, and typically below the laboratory detection limit. In the most recent monitoring round, a lower detection limit of 0.001 mg/L was used. A maximum concentration of 0.01 mg/L was detected in this round, in the sample from BH2.

All concentrations are well below the drinking water MAV of 2 mg/L (MoH, 2005).

There are no clear trends in the copper concentrations which might be related to the remediation works. There were two peaks in BH5 (July 2005 and January 2007) which were elevated compared to other results from that well. In addition, there appeared to be an increase in concentrations in BH2 from late 2007 to late 2008. However, recent concentrations in these monitoring wells have been much lower.

There has been concern expressed that the use of copper compounds as a reagent in the MCD process has resulted in unacceptable contamination of both the soil and groundwater. However, it would appear that the copper is not leaching sufficiently for it to be of concern, at least as measured in the various wells. It is of concern that the detection limit used or most monitoring events was not clearly below the PETC, leaving some uncertainty. However, if greater dilution was allowed, copper is not a concern for the marine environment in the monitored wells.

Subject to measurements of groundwater quality within a better distribution of wells across the site (see Section 1.16.2), copper does not appear to be of concern.

Other Metals

Only limited data are available for the other metals which have PETC, as analyses were only undertaken annually for these elements. PETC exist for chromium (0.44 mg/L), mercury (0.00004 mg/L), selenium (0.5 mg/L) and zinc (0.24 mg/L). Drinking-water MAVs (MoH, 2008) exist for chromium (0.05 mg/L) mercury (0.0007 mg/L) and selenium (0.01 mg/L). The aesthetic guideline value for zinc is 1.5 mg/L.

The chromium concentrations were well below the PETC and MAV for both on and off-site wells. Concentrations were below laboratory detection limits (0.0005 – 0.001 mg/L) except for a single detectable concentration of 0.0022 mg/L at 13 Tahi Street in the January 2007 monitoring round (most recent).

Mercury was detected at concentrations above the PETC of 0.00004 mg/L in the May 2005 monitoring round in BH3 (0.00012 mg/L) and BH4 (0.0177 mg/L). No concentrations of mercury have been detected above the laboratory detection limit of 0.00008 mg/L in recent monitoring rounds. However, the detection limit is above the PETC of 0.00004 mg/L and compliance with this criterion cannot therefore be determined. It is unlikely that there will be significant exceedances of the PETC for mercury as there have been only low concentrations detected in soil on the site.

No selenium concentrations were detected above the PETC or MAV as concentrations were all below the laboratory detection limits (0.001 – 0.005 mg/L).

All zinc concentrations were below the PETC of 0.24 mg/L except for a concentration of 0.63 mg/L which was detected in the well at 36 Tahi Street, although this is below the drinking-water GV. It is unlikely that this relates to site derived contamination as the zinc concentrations in on-site wells were much lower (all less than 0.1 mg/L).

These metals will not be considered further.

7.8.3 Nutrient Monitoring

As noted earlier, there were no PETCs derived for nutrients, as there was no appreciation originally that nutrients could become a particular concern. Results for various compounds analysed for are discussed below.

Nitrate-nitrogen

Nitrate-nitrogen concentrations in groundwater increased significantly from about mid 2007, both in on-site wells and in 13 Tahi Street. Recent concentrations have been lower, but still typically elevated relative to background concentrations.

The on-site wells with the most elevated concentrations have been BH2, BH5 and BH9. The nitrate-nitrogen concentrations in BH2 and BH5 have ranged up to about 300 and 400 mg/L, respectively (in mid to late 2007). More recently concentrations in these wells have been below 100 mg/L. BH9 had peak concentrations of about 200 mg/L and is also currently running at less than 100 mg/L.

Nitrate is both a nutrient and a toxicant, if at high enough concentrations. If a PETC had been derived in a similar manner to the other compounds, but using the NIWA (2002) recommended aquatic guideline instead of the ANZECC (2000) value, then a PETC of 730 mg/L would have been set. Current and past concentrations comply with this value.

The nitrate-nitrogen concentrations in 13 Tahi Street also peaked in 2007 (at about 50 mg/L), with recent concentrations at approximately 10 mg/L or less. In the November 2008 monitoring round, data from other off-site wells showed concentrations above background levels (about 3 mg/L) in most of the downgradient residents bores. Concentrations in these bores typically ranged from about 5 to 10 mg/L. Only one off-site well exceeded the drinking water MAV of 11.3 mg/L (MoH, 2005) in that monitoring round, with a concentration of 13 mg/L detected at 26 Tahi Street.

CH2M HILL were of the opinion that the nitrate detected at 26 Tahi Street was likely to have come from a source other than the site, based on their conceptual groundwater model of little southerly component to the groundwater flow. A leaking sewer or old septic tank was suggested as the source. PDP (2007), on the other hand, considered the result was supportive of a southerly component in the off-site groundwater flow.

Ammonia-nitrogen

Ammonia-nitrogen concentrations in groundwater also increased significantly during the remediation works, although the pattern is not consistent between wells. Recent concentrations have been lower, but still typically elevated relative to background concentrations.

The on-site wells with the most elevated concentrations have been BH1, BH2 and BH5. The ammonia-nitrogen concentrations in BH2 and BH5 have ranged up to about 850 mg/L, although at different times (November 2006 and May 2008 respectively). More recently concentrations in BH2 have been below 100 mg/L and in BH5 have been approximately 10 mg/L. BH1 had peak concentrations of about 600 mg/L in May 2008 and is currently running at less than 200 mg/L. Concentrations in BH9 peaked at about 70 mg/L in late 2006 and are currently less than 10 mg/L.

It may be that concentrations will now stay lower than they have in the past, but given concentrations were high within the last year this is not certain. Further monitoring is required to confirm the apparent reducing trend.

As noted earlier, CH2M HILL inappropriately compared incorrect aquatic guideline values without considering dilution. If 100-fold dilution is taken into account resulting in a criterion of 50 mg/L, only BH1 and BH2 would exceed by up to about three times. If dilution of 1000-fold was used, all recent results (and most results throughout the remediation) would comply. There could be toxic effects at the discharge points where little dilution has had a chance to occur.

The ammoniacal-nitrogen concentrations in 13 Tahi Street have been consistently less than 1 mg/L. The November 2008 monitoring round confirmed similar conditions in other off-site wells, with a maximum concentration of 0.16 mg/L detected.

No off-site wells have exceeded the aesthetic drinking water GV of 1.5 mg/L (MoH, 2008).

Dissolved reactive phosphorus

Dissolved reactive phosphorus concentrations have increased significantly in a number of the on-site wells. The peak concentration for BH2 (approximately 150 mg/L) was detected in the May 2008 monitoring round. The last two results for BH2 have been about 15 mg/L. BH5 has also exhibited elevated peaks, with the maximum phosphorus concentration in this well of about 115 mg/L detected in July 2005. The next highest result in BH5 was 55 mg/L in July 2007, with recent results around 2 mg/L. Phosphorus concentrations in BH1 and BH9 have typically been less than 1 mg/L, other than a peak of about 5 mg/L in BH9 in October 2006.

Phosphorus concentrations in the bore at 13 Tahi Street have been much lower than on-site concentrations, with a maximum of about 0.15 mg/L. In the November 2008 monitoring round, the phosphorus concentrations in other off-site wells were even lower, with a maximum of about 0.06 mg/L.

PDP (2007) compared the results with an ANZECC (2000) marine trigger value for Southeast Australia of 0.01 mg/L. In the interim, until better values are developed, Southeast Australia is considered to be similar to New Zealand conditions. Phosphorus is not a particular concern for on-site wells other than close to discharge points on the foreshores, i.e. in wells BH1, BHD, BH3 and BH5. If a 100‑fold dilution was applied (groundwater trigger of 1 mg/L), recent concentration in BH5 and BH1 are at similar or slightly higher concentrations. November 2008 results on wells near the foreshores, BH3 and BHD, were much lower than BH1 and BH5, at 0.024 and 0.075 mg/L, respectively. The results in the foreshore wells suggest that while phosphorus will be contributing to local effects at discharge points, there is no particular concern after dilution.

Algal growth

Algal growth on the eastern and western foreshores is not formally monitored but it has been mentioned in various reports, particularly the annual TDC biota and sediment monitoring reports. It appears to be a direct response to groundwater with excessive nitrate and phosphorus being discharged on the foreshores following leaching from treated fines. As noted earlier, treated fines are placed close to both the eastern and western foreshores.

Photographs in the TDC reports suggest it is present as a light discontinuous covering over much of the eastern beach and as a heavier growth along a seepage line and locally within the “swale” channel on the western foreshore. Other than the photographs, it would appear that there has been no systematic survey of the extent or effects of the algae. CH2M HILL (2007) noted the potential for effects on local water quality and biodiversity from excessive algal growth. However, in reviewing the comments in various reports, including the TDC reports, no sense has been gained of whether the algae extent is increasing, whether it varies seasonally or whether it is having an effect on the marine biota. No particular alarm has been expressed in the reports reviewed.

Mapping the extent of the algae at quarterly intervals for a year and then annually would assist determining whether it is increasing and therefore whether it could become an unacceptable problem.

7.8.4 CH2M HILL Seep Sampling

CH2M HILL (2007) collected water from two seeps discharging onto the eastern beach by excavating a small pit along a visible seep line. An attempt was also made to collect a seep from the western foreshore but this was unsuccessful. Monitoring of pH and redox potential indicated the water was similar to groundwater although monitoring of electrical conductivity suggested a seawater influence. CH2M HILL concluded the water was indeed groundwater, although impacted by intruding marine water. Given seawater was probably still draining from the sediments at the time of collection, it is likely that these and any similar samples will be at least partially diluted by seawater. Comparing the electrical conductivity of the samples with the electrical conductivity of seawater samples collected during the same investigation, one sample could have been diluted as much as 50% by seawater and the other sample appears to be about 90% seawater.

One sample exceeded the ANZECC (2000) low reliability guideline for DDT (0.0000004 mg/L) by 6000 times and the other by 92 times, indicating the potential for effects. Similarly, dieldrin exceeded the low reliability guideline by 73 and 84 times in the two samples. However, both samples had high turbidity and the results may have been affected by DDT attached to the sediment.

When compared with results from BH1 and BHD, some analytes had similar concentrations, some analytes had higher concentrations (particularly DDX in one seep, suggesting a sediment source) and some very much lower. Given the probable low quality of the samples the seep sampling was of limited value.

It would be useful to repeat the seep sampling exercise on both foreshores in an attempt to obtain a better sense of concentrations in the water discharging on the beach. However, care would have to be taken that the samples were more representative of the groundwater, and not contaminated with beach sediments. Rather than simply excavating holes to sample from, temporary standpipes would likely provide better quality samples. Decanting the samples and centrifuging in the lab would also be necessary.

7.9 Hydrogeological Uncertainties

There are a number of uncertainties and gaps in the information used to develop the hydrogeological model. These are presented in . Further comment is made on the major uncertainty with respect to groundwater levels and flow directions below.

It would be expected that the pattern of groundwater levels would mirror the shape of the peninsula or, more specifically, the shape of the groundwater seepage line around the peninsula. The contours that have been presented all broadly match this pattern; however, there is still a significant difference between the interpretations which would influence the migration path of contaminants from the site. There is also some uncertainty relating to the rainfall, tidal and seasonal influences on groundwater and how their transient effects interact with timings at which readings were taken.

CH2M HILL’s contours are based on readings taken at various times on the 23rd and 24th May 2007 after an earlier rain event on the 23rd May (the original intention of taking all readings within a short period of time on the 22nd May was achieved for on-site wells but not for off-site/residential wells). CH2M HILL acknowledges that as the readings were taken over a prolonged period of time after the rainfall event, some tidal variation may have occurred between readings. It is also possible that recession effects following the rainfall event could have affected the readings. The readings therefore do not necessarily represent a “snapshot” in time and the derived flow directions could be at variance with a true snapshot.²¹

Some indication of the potential effects of tidal variation can be obtained from the figures presented by CH2M HILL, but without knowing the times at which the readings were taken and the corresponding tide levels, the effect cannot be more accurately determined.

The impact of groundwater level recession on groundwater levels can also not be determined from the data available, although CH2M HILL note rainfall responses of 0.1 m in most wells after “a few hours” and 1 m in BH5A after 1.5 hours. The latter figure appears suspicious as this would require approximately 300 mm of rainfall with 100% of this going to recharge in the rainfall event (assuming a porosity of around 30%) to create this rise. Alternative explanations such as the leakage of runoff into the well need to be ruled out. If effects of rainfall are only 0.1 m then this would not significantly alter the groundwater contour pattern, but this cannot be confirmed.

There is also uncertainty regarding groundwater levels to the north of the site. The on-site wells provide a good cross-section through the peninsula and the residential wells provide a good indication of groundwater levels to the south of the site. However, there is an absence of data in the centre and to the north of the site. The potential locations of additional wells are suggested in Section 1.16.2.

PDP (2007) re-contoured the CH2M HILL data with the omission of the off-site wells in order to utilise only those wells for which measurements were taken before the rainfall event. However, whilst levels for these wells were recorded prior to the rainfall event the values presented in the table and those that have been used for the re-contouring were actually recorded at some time after the rainfall event. A second contour plan presented by PDP is based on groundwater levels recorded over two days when no rainfall occurred. The contours also incorporate data from two wells to the north of the site, which help to confirm levels and contour pattern in the northern part of the site. It is not clear from the PDP report if the levels were recorded at the same time during the tidal cycle on each day and therefore what the potential effect of tidal variations may be.

It is worth noting that the groundwater levels recorded by PDP are generally higher than those recorded by CH2M HILL, possibly reflecting seasonal differences, with the higher levels coming out of winter and the lower levels in the lead up to winter. It is also worth noting that in the PDP contour map, the groundwater level in BH9A (on the southern boundary of FCC West) appears to be similar to that recorded previously, whereas most other bores show higher levels than previously recorded. This leads to a slightly unusual contour pattern in this area. It is unclear why this is the case. If this point were ignored then it would be possible to interpret the contours in a way which is a little closer to the CH2M HILL contours but nowhere near as extreme and with a reduced, but still significant, southerly component of groundwater flow.

All of the reviewed interpreted contours are reasonable interpretations of the data and may in fact reflect the natural variation in the shape of the groundwater table under different conditions. This along with uncertainties associated with the data – tidal influence, rainfall influence, reliability of BH9A – means that southerly component of groundwater flow cannot be reasonably ruled out. This could potentially result in contamination reaching groundwater wells to the south of the site.

It should be noted that the remediation does not appear to have had a significant impact on the groundwater flow pattern (although note the lack of data in the centre of the site) and therefore any wells which showed no detection of the contaminants of concern prior to remediation (i.e. were outside the migration pathway) are unlikely to show detections now. The main issue is therefore restricted to any increases in concentrations of contaminants of concern as a result of the remediation activities rather than original level of groundwater contamination.

Unfortunately, the data reviewed contains pre-remediation results for only four off-site wells (13, 17A, 26, and 36 Tahi Street) three of which (all bar 17A) show detects for dieldrin and/or DDE. This suggests that there is indeed a significant component of southward flow (which may be variable dependant on seasonal and/or tidal influences). Of the off-site wells, only 13 Tahi Street shows any response to the remediation activities, although effects in the other wells may be delayed dependent on travel times.

Off-site wells currently show concentration of contaminants of concern above the PETCs and ammonia and nitrate concentrations above the drinking water guidelines. The risk is therefore that concentrations of these will increase further as a result of remediation, although the risk is academic if these wells are not used for drinking water.

7.10 Discussion

The groundwater under the site has residual contamination which will remain for an extended period unless some form of groundwater remediation is carried out. However, it is by no means clear that the groundwater contamination is creating an unacceptable risk, or risks that cannot be managed. The main potentially significant contaminants are DDX, dieldrin and nutrients. Copper does not seem to be an issue.

The location of monitoring wells relative to backfilled treated material is helpful in interpreting the monitoring results. As discussed above, PDP (2007) assessed where various types of commercial quality backfill were placed relative to the watertable. The assessment is not certain, given the lack of monitoring wells to accurately define where the watertable is. However, treated fines, mixed treated fines or untreated commercial material were placed below the watertable or within the zone intermittently submerged, in the following locations close to monitoring wells:

• in all the landfill excavation subgrades, within which monitoring wells BH3A and BH5A were installed (although the position of these wells relative to the clay bund is not clear); and

• in subgrades SG7 and SG9, upgradient of monitoring well BH1A.

In addition, mixed treated fines were placed above the watertable in subgrades SG14 and SG16. Monitoring well BH2 is installed on the boundary of these two subgrades.

It is perhaps not surprising that the highest concentrations measured in wells tend to be where treated materials have been placed in the ground at or close to the watertable, where they can provide an ongoing source of groundwater contamination. These wells are not necessarily the most representative of the site, given that they will tend to reflect extreme conditions.

The mass of DDX, ammonia and nitrate in the ground will ensure an ongoing source for some years, possibly decades. The uncertainty of mass flux calculations means this is difficult to predict. Refinement of the groundwater flow directions and a better distribution of wells would clarify the flux.

The remediation appears to have mobilised the OCP contaminants to some degree, although this seems to have settled down over the last year or so. This mobilisation may have been a result of simply being disturbed by the excavation and backfilling or it may be a result of replacing the soil in locations below the watertable with consequential enhanced leaching. It is therefore not possible to judge whether the SACs derived to protect the groundwater underestimated the leaching potential, or whether breaching the assumptions behind the derivation has made the difference i.e. placing commercial quality material below the watertable and FCC East not yet being predominantly paved or covered in buildings.

Certainly, reducing the infiltration capacity on FCC East will reduce the amount of leaching and therefore groundwater contamination. However, the contribution from the material permanently or intermittently under the watertable is probably much greater than from infiltration. Therefore, reducing the infiltration will not have the same major effect that it would have had if there no treated material had been placed below the watertable. Reducing the infiltration will be only a partial “fix”.

It might be that, in time, the DDX and ADL concentrations in groundwater will reduce as the backfill “ages”. This will occur through substantially irreversible adsorption processes within the soil tending to reduce the mobility of these contaminants, but that will be a long-term effect.

There is no reason to suspect that the highly soluble ammonia and nitrate will reduce in mobility over time. The large source will continue to contaminate the groundwater for at least the short to medium term. The relative proportion of nitrate may increase as ammonia converts to nitrate.

As noted above, it is by no means clear that the groundwater contamination is creating an unacceptable risk, or risks that cannot be managed. Considering remediation of groundwater is therefore premature. Remediation should only be considered when an unacceptable risk is confirmed and it cannot be managed in some other way. Remediation of groundwater would be expensive, potentially uncertain and have to continue for many years. The cumulative cost would be at least many hundreds of thousands of dollars.

The potential risks from the residual groundwater contamination are to:

• existing and potential groundwater users;

• the local effects to the foreshore ecosystem; and

• the wider effects on the aquatic ecosystem.

These are discussed separately below.

7.10.1 Groundwater Users

Abstraction of water in small quantities is currently a permitted use under TDC’s Tasman Resource Management Plan. Potential risks arise for two groups of people; current bore owners south of the site and potential future bore owners within FCC West or south of the site. It is assumed that industrial use of groundwater on FCC East will be prohibited through the use of a management plan or within lease agreements executed by TDC as owner and future tenants.

Management plans are less affective for private residential properties. Unless tied in some way to the property title, management plans may not survive transfer of a property to a new owner. The requesting of Land Information Memorandums (LIM) cannot be relied on (although TDC should ensure that appropriate information is on LIMS that are requested). Education of residents may be effective initially, but with time (and ownership changes) may be less effective. A plan change is an alternative, putting restrictions on constructing bores and abstracting water, but is relatively expensive, not completely certain to get through the public process and slow.

However, it is by no means certain that specific controls are required. To a large degree the presence of a reticulated supply means that the incentive to construct a bore does not exist. Existing bores are used for irrigation, for which the water is of an adequate quality. While drinking bore water from an irrigation supply cannot be absolutely excluded, a child drinking from a hose, for example, occasional exposure carries much less risk.

The drinking water MAVs are derived assuming a 15 kg child (generally the critical receptor) drinks 1 L of water every day for an extended period of time, i.e. chronic exposure. Occasionally drinking bore water, even if contaminated, carries a much lower risk in direct proportion to the smaller average daily consumption. The MAV would increase by the same proportion, perhaps 100-fold, if it is assumed occasional consumption averages no more than 10 ml per day i.e. a small glass per fortnight).

At no point within FCC West or off-site is the groundwater contamination measured to date excessive with respect to MAVs. In fact, at worst, the contamination near the southern boundary and off-site to the south is only marginally over MAVs. Further monitoring is required to confirm this, ideally backed up with the installation of more on-site monitoring wells to obtain a better sense of any southerly groundwater flow component. However, off-site wells to the south generally appear to have an adequately low risk, with 13 Tahi Street (and 18 Tahi Street, if it had a bore) having the greatest risk.

The lack of information on most of FCC West means that the risk cannot be assessed. It must be assumed that the water is unsuitable for human consumption until such time as it is demonstrated otherwise. However, the removal of the most contaminated soil should have resulted, or will result in time, in improved the water quality. This assumes there is a groundwater divide prevent movement of contaminated groundwater from east to west. The groundwater is probably suitable for irrigation use and, given the reduction in soil contamination, may be suitable for potable use.

7.10.2 Local Foreshore Effects

Local foreshore effects are occurring with respect to algal growth. In addition, measured concentrations of DDX and ADL, in wells closest to the foreshore and in foreshore seeps, suggest that there will be local toxic effects at seep discharge points and where seepage discharge runs over the foreshore surface at low tide.

With respect to nutrients the effects are worse than the pre-remediation situation, as nutrient enrichment was a direct result of the additives used in the MCD process, made worse by placing the treated soil in less than ideal places. For DDX and ADL compounds the situation is less clear, with some wells showing small increases and other wells showing small decreases compared with the pre-remediation situation. Comparing recent results with monitoring carried out in 1996 by Woodward-Clyde (1996) there appears to be an improvement in a number of wells, but the 1996 samples may have suspect quality through containing sediment. The more recent baseline survey by T&T (2004) suggest there has been a general deterioration of groundwater quality with respect to OCPs in the wells close to the foreshores, suggesting mass flux of these contaminants is larger now than before the remediation. However, given the concentration fluctuations that have occurred in the current wells and the uncertainty of snapshots several years ago, this conclusion is tentative. A longer period of monitoring would make this more certain.

The question is now whether local effects are acceptable. To meet water guideline values at the points of discharge would require many orders of magnitude improvement for the DDX and ADL compounds. However, given the sediment on both foreshores is still contaminated, and is likely to remain so for an extended period of time then attempting to improve the groundwater further is not warranted.

7.10.3 Wider Effects

Groundwater is discharging to both the Mapua Channel and Waimea Inlet where it will receive large dilution. At low tide the discharge will receive less and delayed dilution while a high tide the dilution will be large and near immediate.

The consent PETCs assumed 100-fold dilution (except for DDX which was 1000-fold dilution). A dilution of 100-fold dilution seems unreasonably small, given the large tidal flushing that occurs. Factors of one to several thousand seem more reasonable. At these factors, dilution should be sufficient to ensure marine water guidelines are met.

Close to the two beaches, however, the CH2M HILL (2007) sampling appears to show that insufficient dilution is occurring. However, the sampling is very limited and has some uncertainties as to whether the samples were representative. The worst results were at low tide. This could be a result of sediment being entrained into the water by seepage flows or wave action, as the samples were noted as turbid. A potential confounding effect is that the sampling was carried out during a rainfall event, which may have caused more runoff across the beach and therefore more entraining of sediment.

It is likely that significant dilution will be occurring within a short distance from the shore. However, consideration could be given to more marine sampling both close to the site and further afield to assess the actual concentrations in the water. This sampling would need to be carefully designed to ensure that the objectives were met. If it is found that insufficient dilution is occurring to meet marine water quality guidelines, a decision then has to be made whether remediation is required, or whether to simply monitor the situation. Given the potential cost of remediation, a proper investigation of alternatives would then be required, including, if necessary, bench trials. International literature abounds with descriptions of groundwater remediation projects that failed to meet the objectives while consuming large amounts of money over many years.