Kirtland Air Force Base Jet Fuel Spill
Sentinel Wells, Contingency Planning and
Data Gaps in RCRA Facility Investigation
Prepared for Citizen Action New Mexico by
David B. McCoy, Esq., Executive Director
April 29, 2014
Numerous uncertainties caused by data gaps exist for contingency planning, modeling and presenting an accurate assessment of the Kirtland Air Force Base Jet Fuel Spill in the Groundwater and Vadose Zone documents produced for the Resource Conservation and Recovery Act (RCRA) Facility Investigation (RFI). Contrary to statements in the Groundwater and Vadose Zone RFIs that no data gaps remain, substantial data gaps do exist. The data gaps encompass: inadequate groundwater monitoring coverage; insufficient knowledge of the hydrogeological setting; accurate values for hydraulic conductivity; accurate values for head boundaries; future groundwater pumping rates and well location. All of these data are important for model validation for arrival times of Ethylene Dibromide (EDB) at Albuquerque’s municipal water supply wells.
1) Sentinel Wells Are Needed at Albuquerque Municipal Drinking Water Wells for Early Warning of Arrival of Ethylene Dibromide (EDB) Contamination.
Kirtland Air Force Base spilled an estimated 24,000,000 gallons of jet fuel and aviation gas into Albuquerque’s drinking water aquifer. (NMED William Moats 2012):
“Officials previously estimated the spill from Kirtland Air Force Base to be about 8 million gallons. But state geologist William Moats, who made the original calculations, recently estimated the spill could be up to three times larger. By comparison, the Exxon Valdez spilled 11 million gallons of crude oil when it ran aground in Prince William Sound in Alaska in 1989.”
The likelihood of EDB reaching drinking water production wells “remains high.” According to the Environmental Protection Agency (EPA), it has no data set for validation of its model prediction that it would be 30 years (p. 65, 66) before transport of EDB to Ridgecrest 5.
In November 2013 CH2MHill (Hill) prepared a contingency plan for the arrival of the plume of the potent disease causing carcinogen, Ethylene Dibromide (EDB), to Albuquerque’s Ridgecrest municipal supply wells, the Veterans Administration Hospital well and the Kirtland well #3.
The CH2MHill contingency plan is not provided on the Water Utility Authority or New Mexico Environment Department (NMED) websites. The report was prepared for the Air Force Civil Engineering Center (AFCEC) but is also not shown in the KAFB document index for public records. Citizen Action obtained the report as a result of a public records request.
The contingencies evaluated in the Hill/AFCEC plan were as follows (p.4-1):
- Wellhead treatment of affected well;
- Contaminant reduction through blending of water;
- Installation of replacement production well away from contamination areas;
- Additional surface water diversion; and
- System operation modification for the Water Authority distribution system.
Kirtland officers boast repeatedly that the Air Force will stop the EDB plume before arrival at the Ridgecrest wells. However, no jet fuel or EDB has been removed from the groundwater since Kirtland became aware of leaking at the fuels facility Pump house in 1992. The Hill contingency plan proposes that the Albuquerque Water Utility Authority consider allowing the EDB to enter Albuquerque’s drinking water and then be blended down to the allowable drinking water limit set by the Environmental Protection Agency (EPA). After an angry response from the Water Authority, Kirtland verbally backed away from its proposal.
The EPA public health goal for EDB in drinking water is zero because EDB is toxic at any level. Long term consumption of EDB at extremely low levels can cause liver, stomach, reproductive system and kidney problems, and may cause an increased risk of cancer, according to the EPA. The EPA drinking water limit for EDB is five (5) times greater than the concentration that some other states will allow in drinking water because of scientific research showing harmful health effects at much lower levels than New Mexico’s adopted EPA drinking water limit of 50 parts per trillion of EDB for drinking water provided to the public. California set its EDB standard five times lower than New Mexico.
In 2012, the ABQ Water Authority adopted a resolution submitted by Citizen Action and sponsored by Councilman Rey Garduno that requested9 that the Air Force install monitoring wells “as close as possible” to the Ridgecrest wells.
The 2014 CH2MHill report likewise recommends the installation of sentinel monitoring wells (p.ES-2):
“The general process for implementing any contingency is to: install and routinely monitor sentinel wells for evidence of [Bulk Fuels Facility] BFF contamination and routinely reevaluate the potential risk to the production wells. If BFF-related contamination is found in a sentinel well, then monitoring frequency is increased and a preferred contingency is selected, planned, and implemented. It is recommended that sentinel wells be installed in upgradient locations for one Kirtland AFB and two Water Authority production wells.” (emphasis supplied).
To the present time, the recommendation of both the Water Authority and the AFCEC for installing sentinel groundwater monitor wells has not been followed. The installation of sentinel wells could not be more urgent given the fact that the EDB is only 800 ft from the VA Supply well, some 4,400 ft from Ridgecrest supply well #5 and 5,000 ft from Ridgecrest #3.
The recent draft models by both the EPA and CH2M Hill for travel time of the EDB plume to the Ridgecrest supply wells set forth arrival times from 20 to 40 years. Citizen Action made a Freedom of Information Act request to obtain the draft EPA report. Rather than furnish it directly to Citizen Action, the EPA report was first provided to a reporter at the Albuquerque Journal. The Journal article failed to consider the glaring data gaps (uncertainties) contained in the EPA Region 6 model that resulted in a badly flawed and unscientific computer model.
2) Kirtland Failure to Inform of Substantial Data Gaps in RCRA Facility Investigation (RFI)
Registered Geologist Robert Gilkeson stated that “The expensive computer models cannot calculate the time for the EDB plume to reach the nearest KAFB supply well (KAFB 3) and the nearest Albuquerque drinking water well (Ridgecrest 5) because there are important data gaps between the plume and the two wells.”
Gilkeson questions why, in all the time since discovery of the contamination plume, sentry wells are not installed at important locations and there has been no tracer dye technology applied to see how rapidly the groundwater is actually moving from one monitoring well to the next. Such technology is available (p. 337):
"The use of tracer dyes is a technically valid and cost-effective method for characterizing contaminant fluxes and hydraulic properties in complex hydrogeologic systems. Dye tracing methods were successfully employed at a site in New Jersey to evaluate the effectiveness of the groundwater containment system and to update the conceptual site model (CSM). The data has driven a reevaluation of the groundwater containment system and CSM, including a review of interim alternative technologies to increase efficiency while a new approach capping the remedial action timeframe at 15 years is tested and implemented."
According to Gilkeson, the draft EPA model listed numerous data gaps that include:
- The location of the leading edge of the EDB plume is not known;
- There is insufficient knowledge of the geologic layers between the EDB plume and wells KAFB 3 and Ridgecrest 5. However , the existing knowledge shows there are fast pathways in northward trending ancient sand and gravel channels of the Rio Grande;
- There is insufficient knowledge of the important property known as hydraulic conductivity (K). A very serious issue is that all of the computer models have used K values that are biased much too low. The low K values provide an obviously incorrect long time for the plume to reach wells KAFB 3 and Ridgecrest 5
Gilkeson, using a formula known as Darcy’s Law, calculates that the EDB plume could reach KAFB 3 within 3 to 5 years and Ridgecrest 5 within 6 to 9 years. These are values very different from the EPA model of 30 years arrival time for Ridgecrest 5 and 40 years for KAFB 3.
The NMED should order the necessary network of monitoring wells to provide early warning for Ridgecrest 5 and other close Albuquerque drinking water wells, says Gilkeson. The hydraulic properties of the geologic layers along the plume pathway must be determined. The installation of monitoring wells and performance of pumping tests is necessary to reduce the uncertainties in the models being used.
Gilkeson points out that at this time there are no monitoring wells along the plume pathway to Ridgecrest 5. The computer models have not erred on the side of, “It’s faster than you think.” Instead, the computer models have used values for hydraulic conductivity that are biased much too low. The result is the prediction of incorrect and much too long times for the EDB plume to reach the drinking water wells. Gilkeson’s conclusion is that the computer models are only providing wild guesses because of insufficient knowledge. “They might as well have used tea leaves or a Ouija board to come up with their predictions.”
A December 23, 2013 Technical Memo prepared for the Water Authority by INTERA, Inc., was obtained by a Citizen Action through a public records request. The Memo comments on the CH2M Hill November contingency plan. Hill postulated an arrival time of 40 years. INTERA stated that Hill’s analysis “does not provide the time when the EDB plume will be captured by the supply wells. Further, “the analysis should not be used to estimate the time when the supply wells will capture the EDB plume.”
Similar to Gilkeson and the EPA Draft model, INTERA found that hydraulic conductivity values used by Hill were too low by a factor of two and limited the value of conclusions that could be drawn. INTERA found that Hill used a flow model based on the Albuquerque basin, a very large area, rather than using discrete measurements from the area between the plume and the Ridgecrest well field. This “significantly underestimated” plume travel times. INTERA recommended recalculation of travel time after obtaining more data for the area between the the EDB plume area and the Ridgecrest wells.
INTERA criticized Hill’s hydraulic conductivity measurements ranging between 1-28 ft per day as in contradiction with Shaw Environmental values of 40 to 200 ft per day. Again INTERA cited the need “to include the measured hydraulic conductivity values in the area of the EDB plume.” It is important to note that the required knowledge of hydraulic conductivity between the plume and the municipal supply wells does not exist.
INTERA concluded that Hill’s scale of the groundwater model “grossly simplified” the groundwater system for the objective of estimating future contamination paths and travel times.
An indication of the foot dragging for characterization of the plume by the Air Force is that in 2010 the NMED informed Kirtland:
“Despite the fact that this release of hazardous constituents was first discovered 10 years ago, [Kirtland] has not completely characterized the Bulk Fuels Facility spill, nor conducted adequate remediation.”
In 2013, the NMED stated:
“Lack of data for hydraulic conductivity prevents designing groundwater containment system for pumping and treating highly contaminated groundwater.”
2) Data Gaps in RCRA Facility Investigation
Kirtland’s RCRA Facility Investigation (RFI: p. 7-2: quantitative evaluation of EDB in soil vapor remains a data gap for ST-106. see here as well) does not comply with EPA requirements cited in the Code of Federal Regulations (40 CFR 270.43) that require full disclosure of all relevant facts and that Kirtland not misrepresent any relevant facts at any time.
EPA requires all necessary data (p. 1 & 4) to support the environmental indicator determinations and for production of a Corrective Measures Study. The nature and extent of contamination requires:
“The Respondent's report shall describe the existing information on the nature and extent of contamination… Identification of areas where additional information is necessary.”
Independent experts, EPA, INTERA and/or the NMED have found:
- data gaps for hydraulic conductivity;
- lack of knowledge for the leading edge of the EDB plume;
- insufficient number of groundwater monitoring wells;
- samples compromised by presence of air bubbles that destroy evidence of volatile organic compounds and;
- shallow well screens that are covered by rising groundwater …
These factors conflict greatly with the statements in the AF RCRA Facility Investigation (RFI). The RFI states that there are no remaining data gaps (pg. ES-7):
The RFI data collected meet the quality requirements set forth in the project-specific Quality Assurance Project Plan (USACE, 2011c), as well as the applicable work plans and letters listed in Section 1, and are sufficient for determining future actions at SWMU SS-111. Analytical sampling, soil boring logs, CSIA and microbe analysis, slug testing, and groundwater modeling were the primary data sources for completing the RFI. The results of this investigation were used to develop a conceptual site model for SWMU SS-111. There are no remaining data gaps for SWMU SS-111 based on the results of the RFI. (Emphasis supplied).
The leading edge of the plume is not identified and there are an insufficient number of groundwater monitoring wells to identify the leading edge of the plume. There is insufficient knowledge of the speed of travel of the EDB plume. The existence of at least a 4,000 ft area between the suspected leading edge of the EDB plume and the Albuquerque municipal wells for the Ridgecrest and Burton well fields is a substantial data gap. Hydraulic conductivity is unknown in the 4,000 ft area and fast pathways may exist. Groundwater sampling is compromised by the existence of air bubbles in samples and many well screens that are submerged beneath the water table. The failure to inform of those data gaps in the RFI is a violation of the RCRA.
The RFI for the Groundwater and the Vadose Zone submitted by Kirtland fail to satisfy data requirements for performance of the Corrective Measures Evaluation and should not be approved by the NMED. Kirtland should comply with requirements for full disclosure of data gaps, immediately install the necessary monitoring and perform the necessary pumping tests to determine: the location of the leading edge of the EDB plume, hydraulic conductivity and to provide early warning for the arrival of EDB contamination at Albuquerque’s water supply wells.
Dave McCoy, Esq. Executive Director
Citizen Action New Mexico
PO Box 4276
Albuquerque, NM 87196