The simple fact is, after three decades, oil still exists underground at the Chestertown hospital site.
Formerly Kent and Queen Anne’s Hospital, the medical center became an affiliate of the University of Maryland Medical System and then a partner, along with Easton’s hospital, as part of UMvShore Regional Health.
The town’s water supply aquifer lies 1,000 feet downhill from the closest part of the oil site. So far, no traces of oil have appeared in the town aquifer.
However, if migration of the oil does come near to the town’s water aquifer, it could signal a well closure along with millions of dollars to set up a new water infrastructure, along with a constellation of problems, from health to property values.
The protection of the town’s water from oil migration since it was discovered in the late 1980s has been done using a “pump and treat” system of monitoring and recovery wells on the site. Considered a successful safeguard, UM Shore Regional Health pays $120,000 a year to keep them running. In 2014 they hired consultants to explore solutions that would clean the site and allow discontinuation of the pump systems.
But how much oil is left? That’s where the complexity begins.
Robert Sipes, has been monitoring the problem since the beginning of his 13-year tenure as the Town Utilities Manager for Chestertown. Using original estimates for the oil spill—160,000 gallons—Sipes calls attention to EPAs own determination that “pump and treat” systems, like the ones we have been using, are only 50% efficient for oil retrieval.
University of Maryland Shore Health consultants claim that the pumps have cleaned 87,000 gallons of oil from the site to date.
By any math, there is an excess of oil left, possibly between 50,000 and 80,000 (some in hard-caked form) still at the site. New solutions have been sought to deal with the rest.
In the Fall of 2014, the Maryland Department of Environment (MDE) approved a pilot test using a soap-based product (Ivy-Sol) to be injected into the site to test the process for its efficiency. The solvent is designed to free the compounded oil so that it can be recovered.
MDE considers the test to be “promising” and while Sipes agrees, he feels that additional safeguards should be implemented, including a site characterization plan to determine where and what levels the remaining oil is and the inclusion of testing for other chemical compounds that have been detected in the area. He warns that simply testing a clean monitoring site after using Ivy-Sol will not be an accurate indication of oil remediation for the total area.
In their interview with the Spy, Sipes, along with former EPA policy analyst David Foster, talk about the current state of the remediation process in light of MDE’s open meeting next Tuesday, July 14th, 6:00 p.m. at Town Hall.
MDE will be announcing and discussing their remediation plan at this meeting.
Dan Menefee says
Job well done on this story! I hope people take the time to see the whole video; there are some important statements towards the end by David Foster. Not mentioned in the video are the concerns of the former MDE official who supervised remediation of the site for 11 years. His comments in a story last year are note worthy as he has extensive scientific knowledge of this site. His comments are crucial and helpful to inform citizens about the history and science of this very complicated issue BEFORE we move forward with more injections.
These were his comments last year:
Fred Keer says:
April 20, 2014 at 5:48 PM
The purpose of these comments is to clarify issues about the site and help promote a meaningful discussion of future remediation steps proposed for the site.
Factual information concerning this site has always been available in the public record since discovery of the accidental release. The specific details of the release could not be determined due to inaccessibility of portions of the system.
A. In excess of 80,000 gallons of number 2 fuel oil (free product) have been extracted from the site.
B. As a general rule no more than 50% of subsurface petroleum can be extracted utilizing primary recovery methods (ground water/water
table skimming). Therefore, a reasonable working estimate of the original release would be 160,000 gallons of fuel oil released to the subsurface. The actual amount could be higher or lower.
C. The town well-field/water supply contains multiple shallow wells.
D. The town well closest to the release site is known to influence the contamination plume. As a pro-active measure, this well was shut down years ago
to help prevent migration of the contamination plume down dip towards the multiple town well-field/water supply.
E. The dissolved phase contamination plume historically migrated down dip/gradient, approximately to the edge of Campus Avenue.
F. The free phase plume historically migrated down dip/gradient to just below Brown Street.
G. The Hospital site lies within the Ground Water Protection Area determined for the town well-field.
H. The town well-field is installed in a paleo stream channel (old stream channel) which trends parallel to Philosophers Terrace
up gradient towards the hospital site. This stream channel is a permeable, preferential pathway.
I. The Hospital/contamination site sits on or near the locality of the headwaters of the paleo stream channel.
No future remediation proposals should be considered for approval without first developing and explaining the site conceptual model postulated for the site. Development of a site conceptual model is standard procedure for all remediation sites. This model must contain the following facts and postulates:
1. A map view of the site annotating all possible source areas and all possible receptor areas. For this site the general source area is the underground storage tank locality and associated contamination plume. The primary, receptor of concern is the town well-field/water supply.
2. Several strike and dip cross sections depicting the detailed stratigraphy (constructed from detailed core data) of the source area including both free phase and dissolved phase contamination plume.
3. Several strike and dip cross sections depicting the town water supply/well-field including well depths and screened/open hole intervals.
4. Cross sections traversing the Hospital site and town well-field/water supply depicting postulated preferential pathways between the two.
The site conceptual model must be the framework for discussing all future remediation proposals and should be refined as new information becomes available.
*Sufficient detailed core data may not be available to refine the existing site conceptual model and evaluate the site for fluid/chemical injection, remediation procedures.
*An accurate estimate of the remaining contamination should be made utilizing core data as well as dissolved groundwater chemical data.
*Prior to permitting/approving any form of subsurface injection of fluids/chemicals, detailed cross-sections of injection well stratigraphy and recovery well stratigraphy must be constructed.
*Injection wells should be designed and constructed for the purpose of injection of fluids into specific formations.
*Chemical makeup of injection fluids must be known and divulged. “Proprietary information” should not be accepted as an excuse to withhold information.
*A robust recovery system and monitoring plan (chemical data and physical data) must be developed to recover and document all injection and recovery of fluids.
*No fluids should be injected without knowing the pathways between injection and recovery wells and monitor wells. These pathways must be determined utilizing detailed core data.
*In order to be legitimate, the monitor well circuit must be separate from the injection/recovery well circuit.
*Contamination (free phase product and dissolved constituent) exists under the Hospital building and for all practical purposes remains inaccessible but mobile. This contamination will continue to be a source area.
The purpose of injection of surfactants into formations is to mobilize immiscible fluids (in this case free phase, number 2 heating oil as a contaminant) trapped/partially trapped within the formation. These mobilized fluids/contaminants will move with the artificially established hydraulic gradient and/or natural hydraulic gradient over time. The natural hydraulic gradient is towards the town well-field/water supply. Once artificial hydraulic control ceases, the natural gradient will control the migration of contaminants.
Given the fact that the contamination plume has been documented to endanger the town’s well-field/water supply, this site seems to be a questionable choice to utilize and/or experiment with known or unknown, surfactant fluid injection, remediation procedures.
Remediation of groundwater petroleum contaminated sites is a difficult, demanding, complicated and expensive undertaking. In my opinion, and as is documented in the public record, the efforts put forth by the responsible party to date have been effective although information put forth to the public could have been better. This is not an easy site to remediate concerning the costs and potential impacts. It is not an easy site to close out.
Patti Willis says
Factual Corrections:
1. While unrecorded at the time, the original oil leak in 1987 was estimated to be 100,000 gallons, based upon the size of the tank and the amount of refueling which had been done prior to the leak. Of the 100,000 gallons, approximately 85,000 gallons have been recovered.
2. The actual annual cost of operating, gauging and sampling the pump and treat system is $120,000 every year.
David Foster says
Although obviously no one was measuring the oil as it leaked out, let me explain precisely why we believe it was well in excess of 100,000 gallons (a figure that has never been verified by MDE).
* First, EPA models suggest that traditional pump and treat systems, such as those employed on this site, normally remove only about 50% of the oil spilled from a leaking underground storage tank.
* Second, the Hospital and its consultants claim to have already removed 85,000 gallons.
* Therefore, if 85,000 represents 50% of the original oil in the ground then the original amount must have been approximately 2 x 85,000 gallons or 170,000 gallons.
* Furthermore, if 170,000 gallons were originally spilled and 50% of those gallons have now been removed, then there must be approximately 85,000 gallons still remaining under the ground and just up the hill from Chestertown’s only drinking water supply.
As this one example raised by Ms. Willis, Communications Director for Shore Regional Health, indicates there is still tremendous uncertainty regarding the quantity of the oil as well as the precise location and mobility of that oil. The subsoil that lies under the ground just downhill from the Hospital can be described as resembling a marble cake, only the dark places in that “cake” are not chocolate but sand and gravel now coated with oil and no one can be certain precisely where those pockets of oil remain. Unfortunately, with this uncertainty comes risk and the real question is who should bear the burden of that risk?
Finally, whether the annual cost of the pump and treat system currently employed is actually $120,000 rather than the $50,000 we originally heard, this is still a very small price compared to the millions of dollars it may cost to replace our wells plus the potential adverse impact on property values in Chestertown. None-the-less, we welcome the new information and we look forward to reviewing the supporting data for that cost estimate, data that has been repeatedly requested by the citizens of Chestertown.
Fred Keer says
Factual Correction: When the spill was discovered the estimete was in the 5,000 to 10,000 gallon range. Nothing more than a guess. Initial spill estimates are notoriously wrong and almost always grossly understated. Secondly, if the Hospital was in posession of a 100,000 gallon estimate it was not transmiteed to the MDE which would have been considered withholding important information necessary for an effectiive remediation design. Does the Hospital staff want to admit this? Thirdly, the only meaningful method to determine the amount of a sub-serface spill is to install sufficient, appropriatley screened mononitor wells, record recovery amounts and apply knowledge of the type of recovery system installed. No credable hydro=geologist would claim, at best, moree than a 50 percent recovery rate for a pump and treat skimming system. Finally, as noted above, the 100,000 gallon estimate was not recorded. Think about it.