Money Talks: Utilizing the Power of Statistics for Remediation Savings

In the world of site cleanup and remediation, the proper use of investigation data and statistical analysis can make all the difference between a huge profit and a ballooning remediation budget. It’s not just about crunching numbers—it’s about maximizing returns, minimizing costs, and saving valuable time. That’s where Fleming Lee Shue (FLS) steps in. We take your data and transform it into a powerful tool, tailored specifically to your needs, rather than forcing it into a rigid regulatory framework. But here’s the secret: it’s not just about having good, validated data—it’s about making sense of it all and using it to your utmost advantage. Let us show you how FLS unlocks the true potential of your data, paving the way to cleanup success. Overall, the environmental field is the perfect application for advanced statistical analysis, but it is infrequently utilized. This is unfortunate because there is a myriad of valuable information that is being misinterpreted, underutilized, or wasted entirely due to a lack of aptitude with statistical concepts. Below are a couple of quick tips/recommendations that can assist you in harnessing the benefits of statistical analysis.
Use All the Data
Examining the importance of the entire dataset to accurately detail site conditions and variations, is critical to formulating cleanup and management plans and is far more effective than just simply comparing results to a standard This is often as far as many consultants will take their analyses. However, understanding the entire dataset and its relationship to the site through a statistical lens allows formulation of critical questions about remediation, management, and negotiation. This information leads to better decisions and comprehensive remedial strategies.  
Prove Your Point
While making your case to a client, mediator, or regulator, it is necessary to know the strength and weakness of your position. Statistics serve tremendously in this effort and let you know ahead of time the strength and limitations of your argument or that of the other party. Statistics foster more informed decisions, allow testing of conceptual models, and provides meaningful answers.  Environmental work is marred and influenced by a multitude of assumptions and prejudices that ultimately can lead to more costs. Statistics empower environmental scientists to cut right through the noise and get to the truth. Results supported by statistical analysis make hard data much more difficult to refute and eliminates the subjectivity and bias that can have your project spinning its wheels. Statistics reveal the evidence.
Really Understand What’s Happing at Your Site | Get Better Results with Better Sampling at Less Cost
Proper use of statistics can reduce your costs by utilizing them during the development of sampling plans! A statistically backed sampling plan can more accurately characterize site conditions and set up the team from the start to obtain more useful and more reliable data. In the world of NYC remediation, million-dollar decisions happen every day. Having reliable data helps you move the needle and ensure your project is moving in the right direction. For example, in one famous well-documented study, a great amount of information was collected, by judgement and convenience, only to have leading statisticians find the conclusions invalid and that better results could have been achieved with far fewer samples had the subjects been properly selected (Eriksen 1999, Fienberg 1990).[1] Similarly, understanding your Site is more than just collecting many samples. Utilizing statistics to develop an advanced targeted sampling plan will help you gain useful data to inform important decisions. And the difference can be millions.
Understanding the Data You Really Have and Navigating Bias
Data bias exists in many forms and can unintentionally skew your perception of a site’s condition. Some types of collection can leave measurements heavily influenced by prior measurements. For example, samples collected too close together are not independent of one another and can result sin bias and loss of information. Similar to carpentry, where you might choose to measure your next cut off of a previously cut board instead of your tape measure. If you do this continually, small measurement errors compound over time and can lead to a big discrepancy. In the environmental world, this might mean you may not have as much useful data as you think, despite having a vast array of samples and data, —and, even worse, the data you have may be seriously biased. It follows, that one has to be careful about analyzing and interpreting it. FLS has in-depth experience utilizing statistics to evaluate data sets and identify inherent bias and make you more certain about the next step.
Challenge the Conventional Wisdom
Conventional wisdom often injects itself into understanding site data. For example, heavier-than-water contaminants, like Dense Non-Aqueous Phase Liquids (DNAPLs), sink through the aquifer and settle on tight soils and lighter-than-water contaminants, Light Non-Aqueous Phase Liquids (LNAPLs), tend to float on the water table. In general, these are simple and true rules-of-thumb. However, taking purely on face value without any analyses to back them up   seriously undermines a remedial effort and can lead one entirely down the wrong path. In one instance, the thinking was that the DNAPL sank and rested atop a confining layer at depth based on earlier sampling. FLS, through proper statistical analysis, found that most DNAPL had not sunk to the lowest clay layer as had been previously thought, but was instead spread throughout and adsorbed to the soil granules well above the depth of the lower clay unit through capillary action. FLS was then able to apply this data to properly position injection wells for effective in situ treatment. Under the original misconception of DNAPL location, the target contamination would likely have been missed or ineffectively treated leading to potential rebound effects, increased remedial activities and obviously increased costs. In both cases, it was a proper statistical approach that led to these successful outcomes.In another case, a petroleum spill, the consultant based their cleanup efforts on observations and instrument reading to determine the limits of contamination leading to an excavation to 30 feet. FLS came aboard this failing project and proposed some limited sampling. FLS then re-evaluated all the data statistically and found that the contamination was limited to four feet, not 30 feet! FLS re-designed the remedial plan that led to successful remediation and closure.  Data gaps and assumptions in the sampling plan had grossly overestimated the contaminated area on Site, by nearly 26 feet and had led to extensive/unnecessary excavation throughout half the property. 
Proper Planning Prevents Problems
The “Four P’s” is a tried-and-true axiom. When coupled with proper and meaningful statistical analysis and analytical methods, these are powerful tools in the hands of the right consultant. For example, sampling planning strategies spelled out in many state guidance documents might include:
·        Collect samples in the 0-2-ft interval, at the groundwater interface, and in areas with suspected contamination.
Delineation of contamination on Site is a great goal. On the surface, this makes sense—sample intervals with the highest likelihood of contamination. However, these methods generally prioritize the collection of biased data to the areas of greatest contamination. By following the constrictive sampling guidelines, we are biasing our data to particular intervals and often overestimating contamination mass present on Site more than in actuality. To actually “delineate” contamination across the Site it is necessary to remove pieces of the human factor, which introduce bias. You can do this by incorporating random sampling, which help capture data in an unbiased way from intervals across the Site ignored by sampling guidelines. This develops a more accurate picture of the Site and when implemented in addition to the more “targeted sampling” approach, will actually help teams delineate contamination across the Site. At FLS we believe this is necessary to effectively characterize the Site and develop an effective remedial plan. Overestimating contamination leads to elevated investigative and remedial costs down the line, but more importantly—in increases to the time to complete the project, where costs can compound quickly! This is a great example of how $1,000 of extra samples early on can save you $1,000,000 in unexpected problems further down the line.
Litigation and Negotiation Support
Combining statistical analysis with legal support yields a very powerful combination to help you in negotiations, dealing with regulatory agencies, and refuting unwarranted environmental claims. Whether it is utilizing advanced statistics to contradict other parties’ assertions or conducting petroleum fingerprint analyses to prove that contamination isn’t from your Site, FLS has years of experience successfully supporting negotiations with the aid of proper statistical methods and graphical analysis.Overall, our clients often come to us because they know we will think out of the box and develop personalized solutions to their unique problems. Through the power of statistics we are often able to see road maps to success—that are invisible to others. Let us know how we can help you! Please see our website at or call 212-675-3225. 
[1] Eriksen, J.A. and Steffen, S.A. 1999. Kiss and Tell, Surveying Sex in the Twentieth Century. Harvard University Press, Cambridge, Mass.Fienberg, S. E., Hoaglin, D.C., Kruskal, W.H., Tanur, J.M., Editors. 1990. A Statistical Model, Frederick Mosteller’s Contributions to Statistics, Science, and Public Policy. Springer-Verlag New York, Inc.