Fast-Track ‘Treatment Train’ Remediation of Pesticides, Petroleum Hydrocarbons and VOCs in Groundwater at a Former Industrial Site in East Anglia, by Regenesis
Shortlisted for Brownfield Awards Category 4 - Best Application of Remediation Technologies and
Category 5 - Best Remediation of a Smaller Site
An accelerated, multi-phase and flexible remedial solution was required to meet a tight lease surrender deadline on a former industrial site in East Anglia.
The site was redeveloped to office use in the late 1980s. Before this a timber yard was present, with an incinerator, fuel storage tanks and a sawmill. Due diligence site investigation was carried out by Ramboll to support lease surrender. Ramboll identified legacy impacts to soil and shallow groundwater in an approximate 500m2 area of the on-site car park. A complex range of groundwater contaminants were present including petroleum hydrocarbons, Volatile Organic Compounds (VOCs), and organochlorine pesticides, which were attributed to historical on and offsite activities unrelated to the Client’s office use. DQRA modelling by Ramboll confirmed a risk to the Secondary ‘A’ Aquifer and nearby river.
The Client’s objective was to handover a site with ‘low environmental risk’ considering potential future uses including residential, within a short timeframe of only 6 months. This required a reduction in groundwater contaminant concentrations by >75%. However, the unusual contaminant mix and elevated baseline concentrations posed a challenge. Pump and treat would not have yielded the mass reduction required in the timeframe set, and the hydrocarbons and organo-chlorine pesticides provided conflicting natural attenuation needs.
Pre-remediation Conceptual Site Model (CSM)
Ramboll’s pre-remediation CSM identified:
a potential offsite source of hydrocarbons beyond the northern site boundary. This had been subject to partial remediation by others historically, however, residual contamination was expected to remain as contamination had been identified directly at the site boundary;
the localised presence of benzene, PAHs, pentachlorophenol (PCP), organo-chlorine pesticides, VOCs and TPH in shallow soils;
a shallow (Secondary A) aquifer comprising silty sand and gravel;
elevated groundwater contaminant concentrations in the shallow aquifer including naphthalene (449 ug/l), trimethylbenzene (TMB) (1,600 ug/l), gamma HCH (129 ug/l), dieldrin (50 ug/l), PCP (2,065 ug/l) and TPH (28,000 ug/l);
contaminant concentration plots inferred a contaminant flux generally from north to south based then on a limited number of wells; and
an inconclusive groundwater flow direction in the shallow aquifer, though this was inferred to be to the west / southwest, towards the nearest surface water receptor.
Soilfix (specialist soil and groundwater remediation contractor) was appointed by the Client and worked closely with Ramboll (Client’s consultant, site investigation, DQRA, Remediation Strategy, regulator liaison, verification) and Regenesis (in-situ remediation design and technologies provider, and on-site injection services) to develop a tailored multi-phase remediation solution for the site.
The ‘treatment train’ remediation approach developed by the project team was as follows:
Primary Remediation: Targeted excavation and offsite treatment of grossly contaminated shallow soils to reduce contaminant mass;
Secondary Remediation: In-situ Chemical Oxidation (ISCO). Injection of chemical oxidants PersulfOx® (activated sodium persulfate) and RegenOx® (sodium percarbonate) to rapidly reduce gross concentrations, particularly TPH; and
Tertiary Remediation: Injection of PlumeStop® Liquid Activated CarbonTM to form a reactive groundwater treatment zone for both hydrocarbons and pesticides through sorption and biodegradation processes.
Figure 1: Remediation Site Plan
Targeted Excavation Phase (Primary Remediation)
The first phase involved the excavation of hotspots of impacted soils, both to reduce potential human health risks for a future redevelopment scenario, and with the added benefit of integrating with the groundwater remediation by removing potential ongoing secondary sources of contamination to groundwater. Extensive field screening for VOCs using a photo ionisation detector (PID) was used to ensure the extent of excavation was minimised. Contaminated arisings were taken to an offsite soil treatment centre as a more sustainable alternative to landfilling.
Well Installations and Additional Aquifer Characterisation
To facilitate injection of chemical oxidation reagents, an injection array comprising a grid of wells was installed across the impacted area. This grid of wells had a dual purpose in allowing additional aquifer characterisation and increasing the resolution of the CSM. This then provided the project team with an opportunity to refine and optimise the groundwater treatment design.
Further characterisation works included:
groundwater sampling and analysis to establish contaminant loading and distribution;
groundwater level monitoring, to clarify the groundwater flow direction and inform design of PlumeStop injections;
Hydraulic Conductivity Profiling (HCP) using a Membrane Interface Probe rig to better understand the heterogeneity in aquifer thickness and permeability, which helped to target the later injection campaigns; and,
a proactive localised PlumeStop injection trial by Regenesis. Inspection of boreholes drilled around the trial injection position showed PlumeStop visibly coating the aquifer soils. This gave confidence that direct push injection would effectively distribute the PlumeStop and confirmed a radius of influence for design.
The CSM was updated after these investigation works to account for:
variable spatial distribution of both TPH and pesticides in groundwater;
variable depth, thickness and permeability of the shallow aquifer; and,
a complex groundwater regime with different flux directions on different scales, including a localised shallow groundwater flow indicated towards the north-west.
Figure 2: Hydraulic Conductivity Profiling Readouts Assessed in Real-time on Site
ISCO Phase (Secondary Remediation)
The aim of this phase was primarily to reduce dissolved phase TPH concentrations (which were typically >2 mg/l pre-remediation), so that the subsequent phase of PlumeStop treatment would be effective in targeting the pesticide and VOC contaminants.
The initial chemical oxidation campaign de-sorbed contaminant mass, pulling it into solution for treatment in subsequent campaigns. RegenOx was initially chosen as the most suitable reagent as the application would pose no risk to existing underground infrastructure/services and yielded more favourable oxidation kinetics for the contaminant mix. During the first RegenOx injection round, injection back pressures and daylighting were greater than anticipated. Regenesis switched to the alternative ISCO reagent PersulfOx, which can be more easily concentrated to reduce the required reagent injection volume. RegenOx was still used in the vicinity of a buried electricity cable due to the potential corrosive effects of PersulfOx on certain building materials.
In total, 4,672kg of PersulfOx and 878kg of RegenOx were injected across four successive injection campaigns (110,000 litres of prepared reagents). Undertaking the ISCO events across separate campaigns ensured that the ground beneath the site did not become over pressurised. Furthermore, as the ISCO reactions progress, more contaminant mass becomes partitioned from the soils into the dissolved phase in groundwater.
The ISCO phase proved successful and reduced TPH concentrations by an average of 84%. An unexpected additional success was that pesticide and VOC concentrations were also significantly reduced, particularly pentachlorophenol.
PlumeStop Phase (Tertiary Remediation)
PlumeStop by Regenesis is an injectable substrate composed of very fine particles of activated carbon (1-2µm) suspended in water. Once in the subsurface, the material behaves as a colloidal biomatrix coating the aquifer matrix. This effectively turns the subsurface into an activated carbon filter with an abundant surface area for contaminant sorption, microbial growth and biodegradation.
As the remedial scheme progressed and more information was gained, the project team was able to enhance the CSM resolution and refine the original PlumeStop design. Key design revisions were as follows:
Based on the pre-remediation CSM it was initially proposed to use the PlumeStop primarily to form barriers along the inferred up hydraulic gradient (northern) and down hydraulic gradient (southern/western) boundaries of the area to treat groundwater contaminants migrating into the site (rebound from the expected offsite source), as well as contaminants migrating away from the impacted area towards the surface water receptor offsite. However, the low confidence in both contaminant distribution and groundwater flux direction(s) meant that significant uncertainty existed in the positioning of such ‘upgradient’ and ‘downgradient’ barriers.
The project team therefore dynamically re-designed the injection layout to create a more concentrated PlumeStop ‘corridor’ along the northern site boundary which was considered to be the most likely downgradient boundary for local groundwater flow based on Soilfix’s groundwater level monitoring. This was also where the most significant contamination had been observed in Ramboll’s site investigation (WS16A) and was adjacent to the known historical offsite source to the north.
An evenly spaced injection grid was designed for PlumeStop application across the remainder of the area. This would act as a reactive treatment zone for residual contamination present within the area, and for any contaminants migrating into the area in future. This approach addressed the uncertainty on groundwater flow directions by effectively ‘covering all bases’.
A more concentrated PlumeStop product (PlumeStop S) was injected locally in areas identified to have more elevated contaminant concentrations post-ISCO.
PlumeStop design was augmented with Advanced Oxygen Release Compound (ORC Advanced®) in areas where TPH concentrations remained highest post-ISCO (approximately 1 mg/l). This was to accelerate aerobic degradation of TPH so that the PlumeStop could more effectively target pesticides.
Regenesis undertook the full-scale PlumeStop injection using the direct push technique, injecting between 1.5-3.5m bgl. The HCP profiling completed during the additional characterisation phase enabled the team to effectively target flux zones within the groundwater body. This maximised contact between treatment reagent and contaminant.
In total, 18,000kg of PlumeStop, 1,200kg of PlumeStop S and 1,125kg of ORC-A (180,000 litres of prepared reagents) were injected into 90 direct push injection points across four weeks by Regenesis’ field team.
Figure 3: Preparation and multi-point injection of PlumeStop
Five rounds of post-remediation groundwater sampling were undertaken. The results showed that the project’s remedial target of reducing contaminant concentrations by >75% compared to pre-remediation had been surpassed, with most contaminants having reduced by >95%. Notably, the more stringent original DQRA criteria (although not set as remedial targets) were also achieved for most contaminants.
Figure 4: Concentration Reductions Pre-remediation to Post-remediation
Figure 5: Contaminant concentration plots against time, showing the successive impact of both the ISCO and Plumestop treatment phases on reducing concentrations
Cost Effective and Sustainable Remediation
The multi-phase remediation approach was flexible enough to adapt to an improved understanding of the CSM and was delivered on programme and in budget. Social, environmental and economic benefits were maximised as illustrated in Figure 6. The project demonstrates the wide-ranging positive effects that a carefully designed and well executed small scale, sustainable remedial intervention can deliver.
Figure 6: Demonstrating Cost Effective and Sustainable Remediation
At the outset of the project the remediation strategy was provided to the Environment Agency (EA), which welcomed the voluntary remediation works. A site visit was undertaken by the Local Authority during remediation, and both the EA and the local authority received a copy of the final Verification Report.
No complaints associated with odours were received from local residents, which was a key benefit of the predominantly in-situ remedial approach.
Health and Safety
Regenesis’ experienced field team managed the risks from specialist multi-well injection equipment and chemicals. Soilfix managed the safety of all works as Principal Contractor under the CDM Regulations. Ramboll provided support to the Client acting in the role of Principal Designer including site inspection visits. No incidents or near misses were reported during the works.
The application of multiple adaptable remedial methodologies at this site has delivered several key achievements:
remediation was achieved within the 6-month timeframe set (November 2018 to May 2019);
the techniques used represent a ‘fit for purpose’ balance between well established source excavation methodologies and more innovative in-situ treatment techniques;
utilising multiple technologies in a ‘treatment train’ approach ensured significant mass reduction of contaminants with conflicting natural attenuation parameters;
an estimated groundwater contaminant mass of >2,300g TPH, >360g PCP, >8g Lindane (Gamma-HCH) and >100g TMB was validated as being removed post-remediation; and,
biodegradation, facilitated by Plumestop, will continue to occur over coming years, providing significant prolonged betterment.