|
Posted
17 October 2009
Communities
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.
Comments
Posted
13 October 2009
Getting
Ready for Winter
While
we're working on the last post in the Risk Assessment series,
I thought some photos taken recently showing what we're up
to, might be fun. Here goes!
More
soon!
Comments
Posted
3 October 2009
Risk
Assessment, Part One
This
is a follow-up post to The
Numbers where we provided laboratory analysis results
for a sample of pit waste at a natural gas well site. Risk
assessment is a huge topic so we'll be splitting the post
into two parts. In this part we'll discuss some features of
the site that have to be taken into consideration and use
information on a
table we've created with various screening levels and
other information for the chemicals analyzed by the laboratory
in the sample.
We're
not doing a full assessment of the site; what we're doing
is trying to find is out if a full assessment toward remediation
is necessary.
The
state's DEP Office of Environmental Remediation has several
publications that have been helpful to us. These are all associated
with their program of voluntary remediation. A
Plain Language Guide to Human Health Risk Assessment
is a description of the process of analysis and decision-making.
There is a helpful checklist at the end which is taken from
Appendix A of West
Virginia Voluntary Remediation and Redevelopment Act Guidance
Manual which is a much more technical document written
for remediation specialists. The third element in the publication
mix is the De
Minimis tables which are part of 60 Code of State Regulations
3. These tables provide screening levels for a large number
of chemicals. (The state's and EPA's tables use exponents
such as E-01 or E+02 with concentration, e.g., 3.89E-01. E-01
is equal to X 10-1 and E+02 is equal to X 102.
For the example 3.89E-01, the concentration is then 0.389.
We find this method of presentation to be a pain and much
prefer either a uniform parts per billion presentation or
as we've presented the figures in our table. A scientific
calculator will easily convert positive and negative exponents.)
When
we had analysis done we just asked for metals, chloride and
radium 226 and radium 228. Chloride doesn't have a screening
level, in spite of the fact that there is a secondary Maximum
Contaminant Level for drinking water for chloride and in spite
of the fact that chloride can be toxic to aquatic life, birds
and mammals. Radiological screening levels are a whole other
topic and since the radium in the sample was at an acceptable
concentration we've put that aside.
Three
of the metals that we had analyzed, even though they show
high concentrations, are not considered important in an environmental
assessment -- calcium, magnesium and sodium. In the end, the
assessment has to focus on arsenic, barium, chromium and lead
(cadmium was not able to be detected by the laboratory), though
we believe the high chloride is an important factor.
The
site is a gas well drilled in 2005 to the Marcellus formation
but two other shales were also fractured. Copies of the well
completion report and plat
are available for download. The well has a pad of about 100
by 200 feet with a drop off to the north where there's a steep
slope into a hollow. On a flat below the well is a spring-fed
cistern (about 326 feet from the well according to our GPS).
The pit is between the wellhead and the drop off and is partially
in fill soil.
This
is on the same ridge (but a mile north of us) that we live
on and we have seasonally high groundwater in winter and spring
with two ephemeral springs close to the house. It's possible
that the site we're examining also has a perched aquifer with
groundwater close to the surface. Some our neighbors, until
this year when city water came to the ridge, depended on spring-fed
cisterns (like the one below the well site), all at about
the same elevation but in different areas, and it's believed
they all are fed by the same aquifer. I don't know if everyone
has city water now or if some still depend on spring-fed cisterns.
The family with the cistern below the well site intended to
use that cistern for a vegetable garden -- their drinking
water comes from a shallow well fed by a deeper aquifer. It's
possible that these aquifers are connected.
The
photographs in The
Numbers post give an idea of what we saw. The area bare
of vegetation -- the hot spot -- is also the lowest spot on
the site and has ponding water.
The
well pad is next to a state road with a house opposite. That
house is about 200 feet from the well and pit area and has
a vegetable garden alongside the road. Since there is a residence
so close to the site and since we believe that it's possible
that the pad area might be used as a building lot once the
well structures are gone, we've considered this a residential
site.
We're
also concerned about the ecological effects that the metals
and chloride would have on vegetation and wildlife. We've
seen plenty of deer tracks in the hot spot on the surface
and assume that deer are attracted to the salts and minerals.
In our assessment, since deer are hunted and eaten in this
area, we have a possible crossover with not just ecological
assessment concerns but also human risk concerns.
We'll
be doing a soil assessment and the factors we'll be considering
include what the sample metals concentrations are, what the
typical background soil concentrations are in this state and
then the various screening levels. That will be discussed
in the second part. Here's that link again to the risk
assessment table we've created for this site.
Comments
Posted
3 October 2009
Risk
Assessment, Part Two
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.
Comments
Posted
26 September 2009
The
Numbers
We
did 13 soil tests on the site using Hach chloride test strips.
In one large area, within a perimeter of torn plastic (approximately
15 by 100 feet), we found elevated chloride in 3 locations:
greater than 650 mg/l at S5 (and nearby in two earlier tests),
333 mg/l at S6 and 136 mg/l at S7. This was along a line through
the area of the site where the pit had been during drilling.
We believe the black plastic is pit liner, used to wrap the
solid drilling waste before the contents of the pit were shallowly
buried under a few inches of soil. Pennsylvania requires 18
inches of cover, the Argonne National Laboratory recommends
3 feet of cover, West Virginia has no guidance.
Here
are some photos to show what we saw:
 |
|
This
photo was taken looking eastward with Molly barely visible
in the background. The red circles show the position
of some of the blocks with flags we set out around the
perimeter of exposed black plastic.
The
bare area in the foreground had the highest chloride
concentration.
|
| |
|
|
 |
|
The
black plastic stuck up out of the ground and was thick
-- nearly impossible to tear.
There
are deer tracks in the soil.
|
| |
|
|
 |
|
This
is the same area as the first photo, looking westward.
The S numbers point to small circles, their locations
along a line about 28 feet apart from each other. |
We
took a sample from the same location as S5 but deeper for
the laboratory analysis. The sample for S5 was collected from
the surface, the sample for the laboratory was from 4 to 5
inches below the surface, entirely within the gray colored
soil area that we felt was pit waste (patches of gray soil
were evident in spots within the perimeter of black plastic).
Here's
what the laboratory found for the requested tests:
| |
Concentration
|
CAS
Number
|
| Chloride |
2550
mg/kg
|
16887-00-6
|
| |
| Arsenic |
16
mg/kg
|
7440-38-2
|
| Barium |
203
mg/kg
|
7440-39-3
|
| Cadmium |
Not
Detected
|
7440-43-9
|
| Calcium |
37100
mg/kg
|
7440-70-2
|
| Chromium |
27.9
mg/kg
|
7440-47-3
|
| Lead |
23.4
mg/kg
|
7439-92-1
|
| Magnesium |
6400
mg/kg
|
7439-95-4
|
| Sodium |
1230
mg/kg
|
7440-23-5
|
| |
| Radium
226 |
1.57
pCi/g
|
13982-63-3
|
| Radium
228 |
1.35
pCi/g
|
15262-20-1
|
Note:
EPA considers chromium (CAS number 7440-47-3) as total chromium
(Cr VI to Cr III at 1 to 6 ratio). Others treat the same CAS
number as chromium III only.
Solid
pit waste isn't uniformly homogenous. If we took a sample
from another location it's possible that the results would
be different. For that reason these numbers give an idea but
not the whole picture.
The
chloride concentration was higher than that shown on the operator's
Discharge Monitoring Report for this well (the liquid waste
was landsprayed) which had 2,125 mg/l. Milligrams per liter
is roughly equivalent to milligrams per kilogram (and both
are also called parts per million). We believe the chloride
concentration in this location is still high because the pit
liner is intact at the bottom. Chloride
has the same mobility as water and won't remain in soil,
certainly not for years unless there's something blocking
its movement.
We
were surprised to see the calcium, magnesium and sodium concentrations
were so high. Generally, they are less than chloride (sodium
about half) for liquid samples, such as liquid pit waste tested
at 5
locations by the DEP in the mid-1980s. The high calcium
and magnesium concentrations offset the sodium, giving the
pit waste a moderate SAR (we
wrote about SAR and the
issue of high sodium in What Happened at Fernow).
Five
heavy metals were analyzed. Cadmium could not be detected
above the lab's effective test limit (0.19 mg/kg). A more
precise test method would be required. The remaining four,
arsenic, barium, chromium, and lead, have concentrations that
may or may not be within West Virginia soil background levels.
That's more research we need to do, though we've seen concentrations
of lead at 28 to 42 mg/kg in uncontaminated soil elsewhere
in this state.
The
metal we're focusing on is arsenic. Some states require notification
if arsenic is found in soil above 5 mg/kg (Alabama, Delaware
-- 2 ppm residential, New Mexico and South Carolina). Kentucky's
industrial soil screening level is 0.185 mg/kg (other states
have lower standards, or higher -- especially for residential
areas). A report providing information about state
regulations covering arsenic in soil for 34 states is
available online.
So
arsenic looks like it's a problem. Fortunately, the radium
226 and 228 concentrations are within acceptable levels. Radium
226 is a long-lasting isotope which breaks down to radium
228 and eventually becomes radon.
What
complicates the issue is the state's approach to drill waste.
Well drilling and production are assumed to have a minimal
impact on the surface. After a well is plugged and equipment
is removed, the surface returns to its original state -- to
be used as farmland, revert to forest, or become a home site
(we've seen trailers moved onto a plugged well site after
the tank and equipment were removed).
As
far as we know, no one has tested pit waste solids in this
state and the DEP has no idea actually what is being buried.
If they've known, they've certainly not told anyone.
This
well and pit are about 200 feet from a residence and vegetable
garden, so a bad pit is a problem for that homeowner. The
pit is about 300 feet away from a spring-fed cistern down
in the hollow, so a bad pit is a problem for another homeowner.
What will happen to the site a hundred years from now will
be a problem for the possible future homeowner living on the
site with a garden and children digging in the soil. Somehow
this isn't an issue for the state or operator.
We
don't think the state can ignore pit waste any longer, or
pretend, once it's lightly covered up, the problem has gone
away.
The
next post will be about risk assessment for this site.
Comments
Posted
26 September 2009
Buckeye
Creek
In
late August the pit holding fracture flowback "water"
for natural gas well 47-017-05815 was breached near Sherwood
in Doddridge County (the north central part of the state).
The pit was constructed within feet of Buckeye Creek (the
state has no requirement for a minimum distance between ground
or surface water for pits -- see our Pits
post) so the "water," at least 2500 gallons, went
into the creek.
The
red gelled liquid has had a negative effect on wildlife. People
were told "it was 'just oil' and hadn't killed any fish
and okay to be in" -- kids swim and play in the Creek.
Already, before the spill, a decline in fish and mussels had
been noted by residents and some of the fish had raised nodules
on the skin.
Here
are some photos:
 |
|
Buckeye
Creek was a good place to fish for bass and muskie.
The contamination is plainly visible from fracture flowback
chemicals and formation material (the color may be do
to high iron) from a Marcellus well.
Gels
are created by chemicals which can include diesel fuel
or ethylene glycol, neither of which is good to swim
in.
A
similar
fracture gel release in Pennsylvania caused a fish
kill.
|
| |
|
|
 |
|
A
high chloride concentration is a feature of fracture
flowback but we don't think chloride killed this muskrat
near its den.
High
chloride will kill fish and other aquatic organisms.
|
| |
|
|
 |
|
Two
ducks were unable to fly. |
Louanne
(who furnished these photos and information) has a letter
she wrote to Governor Manchin available online. The last
I've heard, the gunk has been skimmed from the Creek but is
lying in piles beside the Creek.
Comments
Posted
18 September 2009
Collecting
a Sample for Laboratory Analysis
We'd
been considering collecting a sample for laboratory analysis,
if for no other reason than to confirm the chloride tests
we've being doing for soil and water. James
Otton found Hach strips extremely accurate, but our
way of doing things might not achieve the same results.
The
problem was, we didn't have the slightest idea how to go about
collecting a sample or getting it analyzed.
In
July we began working at a well site where drill waste was
either exposed on the surface of the pad or was close to the
surface. This was a well drilled several years ago and, like
others we've examined this year, has a problem pit.
The
state DEP has a downloadable list of certified laboratories
and that was our starting point. We wanted to find out if
there was a problem with NORM since this was a Marcellus well.
Testing for radionuclides narrowed our choices quite a bit;
most labs don't do radiological analysis. (The downloadable
list indicates by categories what each laboratory can analyze.)
I
contacted Pace
Analytical through their website and a representative
quickly got in touch with me.
We
spoke with two laboratories and their questions were similar,
beginning with "What's the name of your business?"
Pace had no problem working with an individual, the other
laboratory wasn't sure they could. They also weren't aware
of a method for testing chloride in soil, which didn't give
us a lot of confidence in them.
Samples
are either "soil" or "water" and the quantity
submitted and how the samples are collected is important.
This
Texas publication gives a lot of good general information.
Our Pace representative told us that we could, for soil, double
bag the sample using baggies or use a mason jar, and that
a quart mason jar would hold more than enough. (If it were
a water sample we would have needed to collect a gallon.)
A sample for metals needs to be kept cool, radiological samples
can be shipped without cooling.
We
used a wide mouth Ball jar to hold the sample we collected
at the well site. Because we also wanted tests for metals,
we used a latex glove as a barrier between the metal lid and
sample. If we'd had one, we could have used a plastic lid.
(A glass container is good for samples to be tested for metals
and organics, a plastic container is good only for samples
to be tested for metals. A metal lid is no problem for a sample
to be tested for organics.) To collect the sample, we used
a stainless steel serving spoon; a plastic spoon would have
worked just as well. (The
Texas publication provides all the information a person
needs for sample collection and is written for the non-scientist.)
Keeping
the sample cool during shipping required a special cooler.
This
FedEx publication provides useful shipping and packing information.
We found a 1 1/2 inch thick Styrofoam shipping cooler and
box on eBay. If you think you'll be shipping samples to a
lab, prepare well in advance. Ask around, perhaps someone
you know already has a shipping cooler or two.
The
price for analysis wasn't as much as we feared, at least for
metals. Both labs quoted $10 for each metal. Metals are calcium,
sodium, iron, barium, arsenic, boron, etc. They all cost the
same to analyze. Chloride was $15 which wasn't bad either.
Metals testing was relatively quick -- two weeks. Radium 226
and radium 228 analysis cost a lot more ($100) and took 30
days.
The
analysis results arrived by email in the form of a 16 page
report which has pages of technical language in the analysis
narrative and quality control sections. To help understand
the report, we've
found this Alaska state publication extremely helpful.
For further information, searching on the web has also been
useful.
The
lab's analysis has been insightful and worth the trouble and
expense. In the next post we'll give the numbers.
Comments
Posted
12 September 2009
40
Years
Molly
and I first met 40 years ago on Labor Day weekend at a party
during freshmen orientation at college. She was wearing jeans
and a luxurious yellow sweater and was willing to crawl under
a desk to say hi to me.
I
was able to remember her first name and the dorm she lived
in (no mean feat for someone who is name-challenged) and called
her the next day from the campus switchboard (I had no idea
where her dorm was). We met shortly after at an on-campus
movie and our friendship blossomed. We played in the rain,
talked a lot or listened to music in my room. Molly introduced
me to Blues and Muddy Waters.
She
left college after second semester and my best friend and
I visited her in Ohio the next summer after she'd had Daria.
She returned to college a year later, living off campus not
far from my mom's. Until she and Daria found an apartment,
they stayed with my mom. One of my mom's dogs had just had
puppies and I think Daria learned how to bark before she learned
to talk.
I
came back from semester abroad and Molly had left the state
with the man she would marry and I didn't see her again until
years later, though we corresponded in the days before email,
the slow way.
In
1977 Molly and Daria came to visit me in Florida for a month
and after they left Molly and I decided to live together.
I moved to Ohio in 1978 and I count the best years of my life
beginning then, though that first meeting in 1969 was the
ending of the dark ages for me.
Moving
to the woods in 1991, after Daria was in college, was the
beginning of our biggest adventure.
Comments
|
