Risk assessment involves categorizing and one of the side-effects of categorizing is creating clusters or communities — those affected and those not, for instance. Risk assessment, though, goes further and arranges communities within a hierarchical order.

This sort of thinking makes sense in a triage situation or an emergency where response has to be quick and effective. We believe that since there is no public discussion about these categories or the way they are constructed, that the process is open to a bias, no matter how “scientific” or “objective” it may be. One of the biases built into the EPA’s (and the state’s) process of risk assessment is the desire not to create a remediation procedure that is too costly to industry or business. Permitting and other functions, either on a national or on a state level, also have to be concerned with the economic cost to industry.

We believe a bias can be shown to be present if the process is not working — if people are getting sick or if the environment is showing degradation. That’s already happening, isn’t it? Witness global warming. Witness the eradication of aquatic life in 30-some miles of Dunkard Creek that wanders through West Virginia and Pennsylvania. Witness the concerns of residents in the coal fields of West Virginia about underground coal slurry injection and their poisoned drinking water.

The communities we’ll be discussing in this post related to risk assessment are the chemicals, their effects, people and the environment.

The choice of whether a chemical is to be categorized as a COPC (Chemical of Possible Concern) is somewhat arbitrary. Again, this works great in a triage situation. But there are chemicals that aren’t directly toxic to people, that we believe should be members of the COPC category — chloride is one. The state includes chloride in its water quality standards but chloride doesn’t appear on the de minimis list of COPC for 60CSR3. Chloride also doesn’t appear in standard EPA risk assessment tables or those produced by NOAA.

Chemicals that do appear on risk assessment tables are divided into those that cause cancer and those that do not. Generally, those that cause cancer have much stricter screening levels. The problem is that there are chemicals such as endocrine disrupters that have effects on humans and wildlife that could jeopardize a species’ survival. In addition, the presence of a chemical on a risk assessment table generally means that there have been accepted scientific studies about the effects of that chemical. A chemical not on the list should not be considered safe.

In risk assessment concern is about a site’s effects on populations, not individuals. If the community isn’t large enough, then concern lessens. We also have the feeling that if the community isn’t important enough, then concern lessens. This means an urban population tends to have a higher importance than a rural population. Class and race also enter the equation.

One way that states lessen the economic effects of remediation is by having a process where a contaminated site can be, either through deed restriction or covenant, used only for industrial purposes. The risk assessment for industrial soils is many times less protective than residential, and that makes sense in some instances. The process, though, makes it easy to rubber stamp a site as industrial and not do anything about it. In many places industrial and residential sites are not carefully delineated — industrial sites have close neighbors who live in houses.

The site we’re looking at is not located where its effects will be felt by a large enough community to make it very important from an official, risk assessment, standpoint. Only a few nearby residents could possibly directly experience the results of soil or groundwater contamination. Again, in a triage situation, such thinking might be appropriate (though racial and class bias would not), but here we believe that the possible effects on individuals warrants closer examination.

The bias that is shown for communities of people in risk assessment is heightened when it comes to environmental screening. Ecological communities are carefully delineated and extremely narrow and include endangered species, old-growth forests, and federal and state parks. Our belief is that if the contamination at a site can be assessed to cause a risk to wildlife or vegetation, then that is a problem, even if the wildlife or vegetation doesn’t have the luxury of being rare or located in a park. We’re especially concerned about the numerous instances we’ve seen at well sites where deer have been attracted to areas that have been contaminated in one way or another. Deer are attracted to salts and minerals present but it’s possible they are also ingesting chemicals that could do them harm, or harm those further up the food chain.

Risk assessment is a tool but the way the tool works is often wanting.

We’ll be referring to the table we have available for download for the screening levels for assessment.

For this state, assessment of soil contamination proceeds in a systematic manner (see the Decision Tree document). The sample’s concentration is compared to background levels — either uncontaminated soil from or near the site or state background levels. If the sample’s concentration is lower than the background level, then there is no problem. If the concentration is higher then the next stage of assessment is made. For our samples, arsenic and lead were higher than the maximum concentrations in the state’s background levels. Barium and chromium were lower so those metals aren’t considered a concern. (Canada is currently creating national guidelines for acceptable concentrations of chemicals in soil. These limitations will be in force, even if the soil background concentration is higher.)

The next step is to compare arsenic and lead concentrations to the soil to groundwater screening levels. The EPA recently changed their screening levels (June 2009) and we’re using the EPA’s current screening levels for this region. The sample’s concentrations for lead and arsenic were both higher than the soil to groundwater screening levels (arsenic is also higher for the state’s soil to groundwater screening level — these for some reason are about 20 times higher than the EPA’s). Typically this means that groundwater needs to be tested for these elements.

The next step is to consider residential screening levels. In this case, the arsenic concentration is still higher than the residential screening level, but lead is lower.

In West Virginia an option where soil screening levels are high for residential but don’t exceed the industrial soil screening level, is a deed restriction, so that the property can only be used for industrial activity and not as a place to live, garden or farm. We don’t believe that is an option here since the well pad is on a half acre of a much larger piece of property. The natural gas operator doesn’t own the property, they lease the mineral rights from someone else who owns the mineral rights. Surface rights were severed from the mineral rights some time ago.

Another concern for all the metals — arsenic, barium, chromium and lead — is how much higher their concentrations are than ecological screening levels. In the case of arsenic the eco-ssl is much higher than the residential soil screening level (but the sample’s concentration is higher still), while for barium, chromium and lead the eco-ssl are much lower than residential screening levels.

Our preliminary assessment comes from the laboratory analysis of one spot within an affected surface area of at least 1500 square feet. The matrix has a high chloride concentration which would enable the transport of chemicals through soil to water beyond the confines of the site. It’s entirely possible that laboratory analysis at other points would alter the preliminary assessment’s chemicals of concern and add others. It’s also possible that the pit liner is no longer intact at the bottom further north of the sample location and migration of pollutants has already occurred.

For these reasons, we believe the site deserves a full assessment by a professional though we believe that remediation options for the operator are limited. We hesitate to make recommendations but believe the operator’s preference of doing nothing is not viable because we believe that the site’s unremediated presence is a danger to those living nearby.