Contaminated Land Reports for Planning - A Peer Review

After 15 years of peer reviewing reports submitted for planning and over 3,500 reports / responses reviewed I like to think I’ve seen a lot. From forged ground gas verification reports (someone went to jail for that one), someone arguing no visual or olfactory evidence of carbon dioxide, ground gas monitoring done by waving a confined space monitor over a trial pit and everything in between. I've seen some of the worst and some of the very best that this industry has to offer. Here’s my round up of the top 10 things that some of our report writers and risk assessors out there should consider when writing their offerings for planning.

Before I start the list, a little about me. I’ve worked for large consultancies since starting in the industry in late 2008, working on everything from small brownfield development sites right up to power station redevelopments, pharmaceutical research sites and former ordnance factories. The usual large consultancy upbringing with more than a few spicy sites investigated / remediated in my time. Whilst doing all of that I’ve also been diligently working away on reviewing reports submitted for planning for up to 6 councils over my career.

I always had a passion for these projects, not for the prestige but perhaps helping the various Local Authorities promote ‘safe development’ in their areas appeals to me, or maybe it's being a bit of a detective, or maybe I’m just argumentative? I’ll let you decide that one. It has given me the opportunity to review large numbers of reports from every echelon of our industry, from small one to two person outfits, right up to the large multi-disciplinary consultancies.

This is the sort of work that you need a particular skillset to do, it takes experience and a certain way of thinking to spend a couple of hours with a report – strip it back to its basics, target the loose threads and pick apart the bones to focus on the key issues that aren't fully developed. The aim is for your client (the Contaminated Land Officer) to read it and get the key points digested do that they can make sound regulatory decisions. Sounds easy, right? Truth is it can be, but when you combine that with the intricacies of how each council area regulates and the position of the planning applicant and their agents (i.e. is it a self-build project having a go, a structural consultant trying their best or an environmental consultancy who knows the trade), the way you word the subsequent commentary on their report needs to be carefully considered.? It’s not just knowing the guidance or reviewing the report, how you word that query to get the request across in a neutral way, protect your client (the council – and arguably those buying the houses if it’s a residential development), and get the answers you need without just specifying a scope of work or providing the answer is also important.

Throughout doing this over the last 15 years, I’ve seen the same recurring themes coming up in my commentary on these reports, with similar issues cropping up time and again – regardless of who the submitting consultant is. This is my top 10 list, and (being a consultant) I should note now that caveats and limitations do apply. This is based on my experience in the areas I’ve worked in – your area may differ, and your mileage may vary etc.

1. Over-summarised and over-simplified PRA

The Preliminary Risk Assessment (PRA) is supposed to form the foundation on which our other risk assessments are built on and is a significant part of informing the subsequent ground investigation - or showing that risks are low. Too often I see over-summarised PRAs that lump all sources and pathways together, only risk assessing receptors rather than individual sources and pathways. I think (or more accurately, I hope) this is done with the aim of simplifying this so that the developers can easier understand the assessments, however, the source-pathway-receptor relationships we’re assessing are rarely that simple. When reviewing a report, if the CSM / PRA are over-summarised and unspecific, it has the unfortunate after-effect of eroding confidence that the site and it's issues have been fully characterised, pushing doubt forwards into the ground investigation and Generic Quantitative Risk Assessment (GQRA). A good PRA should split relevant sources and risk assess against individual receptors and individual pathways, focusing the subsequent ground investigation and GQRA on confirming the sources and pathways of concern. In essence, the PRA should be an easy to digest summary of all the key concerns from a contaminated land perspective. This process works. In a single table you can effectively demonstrate the key issues to the client and / or the regulator in a qualitatively measured way that is directly linked to guidance. In the words of French winemakers: ‘traditions are experiments that worked’, there’s a reason CIRIA C552 from 2001 suggests that PRA is done this way and that the process hasn't changed in over two decades, it works.

2.?????? The Death of the GI Strategy

This is a problem across board it seems. The GI strategy is a useful hold point between the desk study and the groundworks phase to really focus the subsequent ground investigation on what was found in the desk study. It’s where some of the below items in the list could be discussed, designed, focused or even discounted. I’ve only ever reviewed one GI strategy in the 15 years I’ve done this. I understand why this gets dropped a lot of the time, as from a business perspective it’s an easy win to reduce costs and win work – but at what sacrifice I wonder? So many of the reports I see would have really benefitted from this hold point, taking a view on the specific issues needed to be resolved from the ground investigation and taking a pause to get that all important regulatory buy-in to the intrusive work proposed.

3.?????? Cross Installed, Cross Wired GI

If had ￡1 for every cross installed monitoring well I’ve seen in reports coming across my desk I would probably be able to retire at this point. Is this not one of the key things we teach graduates / consultants when they’re taking their fledging steps on site? I grew up with the Southam, Carr and Murray School of Install Design where this concept was hammered into me, one project at a time. It’s understood that there are obviously exceptions to these rules (groundwater rarely respects boundaries after all), but cross installs should be the minority not the majority, right? Getting these wrong pushes doubt forward into the controlled waters and ground gas risk assessments, leaves further explaining to do later, potentially give results that may not be usable or (more worryingly) result in incorrect decision making at risk assessment stage. The question I ask myself is why does this happen? Is this not being taught? Is it the result of tired engineers desperately trying to keep up whilst being thrown headlong into the next investigation without the time to demob the last one? Is it people from other disciplines specifying the install details? Or are we trying to make our monitoring wells too multi-purposed in an attempt to compress costs? Genuinely interested on people's views on this one so let me know your thoughts.

4.?????? Issues of a Gaseous Nature

Ground gas risk assessments I see coming through the peer review system vary significantly more in their content and implementation than almost any other element of contaminated land risk assessment. I think we (as an industry) do far too much ground gas monitoring, much of it on sites or projects (highways projects being an example) that may not even need it in the first place. Steve Wilson 's various presentations cover this infinitely better than I ever could, so I’d recommend watching those whenever you can if you’re coming into the industry (link to the SoBRA EC webinars below). This is something that really should be covered more in desk studies, where the preliminary ground model should be broadly assessed for ground gas generation potential at this early stage. A case can then be made using BS 8576:2013 as to whether monitoring is even needed, which can be re-considered once the ground is broken at ground investigation stage. Suffice to say, once you have a dataset that is full of randomly high flow rates and / or elevated concentrations with no real sources or as a result of poor installations, it’s then much more difficult (but by no means impossible) to explain these away. Worryingly, when some of these risk assessments are challenged with some basic questions in the peer review process, a proportion of the consultancies just state that they’re upgrading and putting ground gas protection measures in rather than just adding more lines of evidence to support their argument. Often there is little attempt to consider the dataset as whole, even when it doesn't match the gas generation potential, most just consider it easier to be conservative and specify protection measures. Once that decision is made, that opens up a whole new chain of questions about how these are scoped, designed, installed and verified. Yes, before this gets pointed out in the comments, there are some Local Authorities who take a much harder stance on ground gas risk assessment than others, and for good reason as most of those have legacy coal mining issues. But generally speaking, it isn’t the Local Authority’s responsibility to fill in the blanks in the risk assessment, do that right (and in accordance with guidance) and there isn’t much a Local Authority can do other than accept the conclusion. It also isn't the Local Authority's role to reconsider the risks and whether protection measures are overly conservative, that remains with the developer and their agents.

5.?????? Mine Gas, a Working Problem

Since the release of Contaminated Land: Applications in Real Environments (CL:AIRE) guidance on mine gas risk assessment in 2021, it feels like the contaminated land discipline is still is catch-up mode on mine gas risk assessment in general. RSK’s recent publication on this issue for the Scottish Government (link below) makes it clear that even in the regulatory space, knowledge on this issue varies from council to council. It’s a hot topic in the industry at the moment and when I have raised this issue and provided some basic information at both my previous and current employers, the phone rang constantly for days afterwards with questions, queries, sites that might need further work etc. In the peer review process, I’ve seen the usual cookie cutter ground gas risk assessments coming through the industry with little or no recognition of the guidance or attempts to just hand wave it away as some far-off problem that doesn’t really affect this site in particular. I’ve had various rebuttals in the last 3 years and “it’s just guidance, not instructions” statements come through with no real recognition that whilst it is the same media in terms of contaminant, the two different sources (soil gas and mine gas) are now assessed in a very different manner. The guidance is clear that Gas Screening Values alone should not be relied upon in this situation, but they're used as a significant crutch specifically for mine gas in almost every report that crosses my desk. I've sat in meetings with previous CLO clients and applicant's agents where I've essentially been told 'well just tell us what needs writing and we'll write it, we already know it won't change our conclusions at all'. Not even getting into why that statement isn't one I would personally put to a regulator, it does show an apparent (and worrying) unwillingness amongst some to even recognise that mine gas is a bit different to soil gas, and that it is usually a potential high risk source. It’s definitely not fully understood and engrained across the industry yet based on what I’m seeing, but since the NHBC published their updated guidance on ground gas in April 2023 it feels like traction is starting to build on this. I’ll be watching this particular one with interest as it develops.

6.?????? Underdeveloped Controlled Waters / Water Environment Risk Assessment

The water environment or ‘controlled waters’ if you’re an England / Wales based consultant - I have the unfortunate issue of having to be bilingual here, being an English based consultant working on an approximate 50:50 ratio of Scotland / England projects. I think this particular one may be localised to the Glasgow-Edinburgh belt, where the prevalence of thick deposits of Glacial Till are pervasive across the area. From what I see coming through the system, it is very clear that the water environment (particularly groundwater) is routinely just discounted at desk study stage, with limited or no water environment sampling and / or analysis completed. Before I get shouted at in the comments section, I’m not saying that this is wholly unreasonable in basic principle, but it needs to be covered with evidence-based discussion pointing out the various lines of evidence that are available to discount these receptors (more on that below). Simply stating ‘Glacial Till = no problem’ won’t cut it if there is no evidence to confirm thickness, composition, extent across the site etc. I’ve even had this justification provided where all exploratory logs on the site record sand from top to bottom, go figure! The problem is, from my recent reviews for other council areas outside of the Glasgow-Edinburgh belt this idea seems to have bled into the whole of southern Scotland, and I'm seeing incredibly limited water environment risk assessment undertaken virtually across the board, even where groundwater is identified in significant quantities in monitoring records.

7.?????? Absence of Evidence is not Evidence of absence

I liken this one to presenting the answer without showing your working. I too often see ascertains that ‘risks are low’ with limited lines of evidence presented to show that to be the case. Admittedly, in some cases the right answer appears to have been reached but if this is effectively an ascertain that is based on a single line of evidence only, it will likely be picked up on review. This was a valuable lesson I personally learnt from Alex Lee back in 2010, who persistently (and at the time, annoyingly) challenged my conclusions and teased out the various lines of evidence that I had considered, but not necessarily written down. It's a lesson I still remember to this day. This is set out quite clearly in LCRM, which states: “The data you collect must be relevant, sufficient, reliable and transparent – you must check the quality of the information. When we assess results for the sites we regulate, we will take into account the entire dataset. We will use a ‘lines of evidence approach’ to make any regulatory decisions”. Put simply, a thoroughly explained and evidence-based argument in accordance with relevant guidance and recognising potential uncertainty is highly unlikely to be challenged. Why then are the vast majority of the queries I submit back in these reviews simply asking for more evidence to back up the conclusions reached? It might take an hour’s more thought, research and writing, but it will save time and the client’s money in the long-run if this is right first time. The old adage of “do it right, do it once”.

8.?????? The Hot spot problem

You’ve got your analytical data back and run it through your relevant data screening tools, and there’s that one annoying lead concentration that just happens to be over your human health assessment criteria – what to do? There seems to be a prevailing idea going around the industry at the moment that simply digging out a 5m by 5m box for example with validation sampling is sufficient to just make that problem go away. This might be ok for a targeted sample taken from a known and localised area, a former backfilled pond for example – but even then, you should be digging out the whole of the pond backfill unless you have a sample population big enough to do stats in that case. Often this is also specified for non-targeted samples with no rhyme or reason as to why that particular sample exceeds the assessment criteria. Before declaring that this is a hot spot, it needs to be established as a hot spot by use of a histogram or a named statistical test. Most of the theory behind this is set out BS 10175:2011 (+A2 2017) , albeit not directly, but further clarified in CL:AIRE 2020 Professional Guidance: Comparing Soil Contamination Data with a Critical Concentration. The whole idea of a sampling grid is that it statistically (via selection of an appropriate grid type and size) increases our chances of identifying potential unexpected areas of contamination. If found, the additional statistical work is needed to prove whether that annoying sample is part of the general sample population (and therefore potentially representative of other, untested locations in between our sampling grid locations) or a statistical outlier (and therefore, a hot spot!). This doesn’t seem to be common knowledge across the reports that cross my desk for peer review, and I comment on this for approximately 60% of the reports I see.

9.?????? Modelling the overuse of Models!

Modelling can be a really powerful and useful tool in our industry, if done right, and always on the understanding that empirical data beats modelling hands down. I learnt that lesson from Mark Sharpe, who recalled a meeting where a consultant swore blind that their model showed a contaminant would take 1,000 years+ to reach the receptor, meanwhile Mark had data from a borehole on his site adjacent to that receptor recording that contaminant there already at elevated concentrations. In a larger consultancy, modelling seems to be saved for those larger sites with significant contamination migration potential, though I’m starting to see a trend in reports of modelling being used for relatively minor (5% over a water quality standard for example) exceeding concentrations for surface water receptors which are hundreds of metres from the site. Why? Why model something this minor instead of using a much simpler source-pathway-receptor justification with your lines of evidence to demonstrate low risks? I’ve even seen modelling being used in place of much simpler methods such as M-BAT assessments. The minute a model is used, this raises a raft of additional questions: what input parameters have you used and why? What compliance point have you selected and why? What are the dimensions of the source area and what is the justification for these? What does the sensitivity analysis show? The accuracy and applicability of models is almost entirely dependent on the all important input parameters, so don't just grab the first value you can find from Freeze & Cherry 1979 when a simple Google search could find geology-specific BGS values from much more recent studies that are orders of magnitude in difference. Even better than this, get some geotechnical and hydrogeological information from the ground investigation and test the parameters of the geology beneath the site specifically. I think as an industry we need to get back to basics on these risk assessment principles rather than over-complicating things, yes, modelling is an important tool – but save it for those sites that really need it.

10.?? Well Decommissioning

The Gorebridge case shows better than almost anything else why decommissioning of monitoring wells (as well as their design – covered above) is an important step in the process. Most of the time this is something that has to be prompted in the review process, but in fairness, most consultancies whose reports I have reviewed are happy to undertake this once that prompt is put to them. The issue comes around verification, which is not actually covered well in guidance. Typically a decommissioning strategy is agreed (following existing Environment Agency or Scottish Environmental Protection Agency guidance), with verification often comprising a letter report detailing the works with photography of wells pre and post decommissioning provided. I personally think it would be good to have some (succinct - its not a difficult concept after all) clarity in guidance on the verification process for monitoring well decommissioning. CL:AIRE 2021 guidance and RSK’s recent publication for the Scottish government on mine gas risks makes the importance of this pretty clear from a ground gas point of view, with existing Environment Agency and Scottish Environmental Protection Agency guidance making this clear from a groundwater protection perspective. This should be a relatively basic recommendation in all ground investigation reports to be honest, but it’s not one I see regularly on reports that cross over my desk. Even where decommissioning is undertaken, many monitoring wells are simply destroyed during construction works and end up not being decommissioned.

The Potential Cost of all this...

Local Authorities like most other public services are under increasing financial pressure, CLOs I have worked with over the last decade and a half have all told me of the various tolls being taken. Whether it is increased workload, the combination of Contaminated Land into Environmental Health (therefore sharing responsibilities between contaminated land and the more regular environmental health duties), the ever reducing budgets, the loss of good EHOs / CLOs and the difficulty in their replacement.

Objectively, having to review and face these reporting problems will be adding to these hardships, either by repeated review or follow-up in house, or (as a number of authorities now do) seeking assistance from private sector consultants to help with the workload. Whilst you may think I should be celebrating this as a private sector consultant myself, I cannot help but have some sympathy with those Local Authorities I work with - who are diligently having to deal with this day in and day out.

Some of this is undoubtedly driven by the market, whereby some developers are relentlessly chasing cheaper prices and not necessarily understanding that they're being offered 'just enough to get it through' services to allow those cheaper costs - and, for the avoidance of doubt, this also includes big multi-disciplinary companies. It really gets the point across that you get what you pay for, if time is of the essence (like it is for almost every other project I work on) then getting a good and comprehensive service really helps save that time at the critical point of the project - mobilisation for the build. Meanwhile, the costs and burden of dealing with these reports is being pushed onto already pressured Local Authorities in further back and forth communications and reviews trying to get to that point where all contaminated land issues are reasonably addressed.

Personally, I hope we get to the stage where this is less of an issue, but for now, it seems that I'll have plenty of work to do on these for years to come.