Guidelines for Vapor Monitoring INTRODUCTION Due to the complex nature of vapor movement through soil, the following guidelines have been developed to determine when vapor monitoring can be used as a method of release detection. These guidelines apply equally to facilities with or without existing contamination. Owners or operators choosing to use vapor monitoring must submit a Vapor Monitoring Plan (VMP) incorporating three primary elements (described in detail later in this document) to the Department for review: 1. a Site Suitability Assessment for the viability of using vapor monitoring; 2. a Monitoring and Sampling Plan that outlines the method of vapor monitoring to be utilized; and 3. a Data Management and Reporting Plan. The Department must approve the VMP before vapor monitoring can be used as a method of release detection. Another method of release detection in accordance with Chapters 62761/762, F.A.C., must be used until the VMP is approved. If the Department does not approve the VMP, vapor monitoring will not be acceptable as a method of release detection. However, a revised VMP can be submitted for review and approval by the Department. Any modifications to the VMP (such as change in the vapor monitoring equipment) shall be submitted to the Department for approval prior to instituting the change. In addition, any variation from the VMP in site conditions or monitoring results must be documented and incorporated into a revised VMP. The revised VMP must be reviewed and approved by the Department and another method of release detection in accordance with Chapters 62761/762, F.A.C., used until the revised VMP is approved. If it is not approved, vapor monitoring will not be acceptable as a method of release detection. The DEP-approved VMP shall become part of the facility's records and be available for inspection within five working days following notification by the Department or locally contracted program. All records collected pursuant to the approved VMP must similarly be made available. I. SITE SUITABILITY ASSESSMENT: For vapor monitoring to be an effective method of release detection, the site must be conducive to such a method. Therefore, the site must be assessed to determine the feasibility of vapor monitoring to detect a release from any portion of the storage tank system for which vapor monitoring is being utilized.
Data must be gathered to assess time-related variations in site conditions. The set of site characteristics data must be sufficient to be statistically valid and taken in time increments no greater than 30 days. On the average, a calendar year's worth of data will be needed. The site suitability assessment must include: 1. a description of the physical properties of the product stored in the storage tank system, including the volatility and specific gravity; 2. a determination that the use of proposed or existing monitoring wells or probes are suitable in consideration of their location, construction, and the soil air permeability between the storage tank system and wells or probes; 3. a determination of the time-related variations in groundwater depth and flow direction. Where applicable, the variation of the differential head between the storage tank system product and the ground water table must be addressed; and 4. a determination of the source, magnitude and time-related variation of background vapors, including the environmental factors influencing these readings. If the following specific site suitability criteria (Monitoring Well or Soil Vapor Probes, Groundwater Characteristics, and Background Vapor Characteristics) cannot be met, then an alternate means must be used to support the claim that vapor monitoring would detect a release within 30 days with a probability of detection/false alarm of 95% / 5%. Some options are: • • • diffusion calculations that address the movement of the volatile constituents through soil, as influenced by soil porosity and moisture content. tracer tests that use a known or marked volatile compound to mimic the movement of the volatile constituents of the product through the soil. empirical evidence that provides practical confirmation that vapor monitoring will detect a release in accordance with Chapters 62-761/762, F.A.C.
A. Monitoring Wells or Soil Vapor Probes General information: The use of soil vapor probes for vapor monitoring instead of wells may reduce the problems of vertical stratification and the associated need for purging. However, the guidelines for site suitability apply equally to probes and wells.
(1) Wells must be constructed in accordance with Rule 62-761.640, F.A.C. Boring logs, etc. must be included as available. (2) The wells should be located within 10 horizontal feet of any potential leak point. The 10 foot distance acts as a radius of detection around each well. Therefore, well spacing must be designed to afford complete vapor detection. (3) The majority of wells shall be placed down gradient of the system. (4) For underground storage tanks or underground integral piping, the well depth should descend no greater than 2 feet below the component being monitored. For aboveground storage tanks, the well depth should descend no greater than 4 feet below the storage tank. Unscreened probes must be installed in clusters at multiple depths. (5) Soil air permeability must be 1 Darcy (1 Darcy = 2.738 feet per day and is a measure of permeability) or greater, as determined by laboratory or field soil permeability test, or be classified as a medium or coarser grain sand in accordance with the Unified Soil Classification System. (6) The wells or probes must not have sheen or free product. B. Groundwater Characteristics
(1) Depth to groundwater should be at least 2' below the lowest point of the storage tank system component being monitored. If groundwater levels are above this point, demonstration must be made to show that the product head is above groundwater head at least once per month. (2) A minimum of 2' of well screening shall always remain above the water table. If unscreened probes are used, the termination point of at least one probe per multiple depth cluster shall be at least 2' above the highest level of the groundwater. (3) The method of determining groundwater flow must be in accordance with standard practices. C. Background Vapor Characteristics
(1) Naturally occurring background vapors must be differentiated from the vapors caused by preexisting contamination.
(2) Time-related variations of background vapors must be assessed. Vapor sources or environmental conditions that interfere with the determination of a release must be incorporated in the Release Detection Response Level. (See Section II.B.) (3) The vapor monitoring method and equipment must be used in a manner that compensates for the inherent limitations of the detection devices. (See Section II.A.) (4) Demonstration must be made to show that any active remediation on site will not interfere with the detection of a new release. II. MONITORING AND SAMPLING PLAN The Monitoring and Sampling Plan shall describe the specific procedures for sampling, the vapor level at which a release may be suspected, and the notification procedures to be followed when the level is reached or exceeded. The Plan shall include the following: A. Method of Sampling There are several instruments that can be used to measure organic vapors in a vapor monitoring well. Vapor detectors such as flame ionization detectors, photo ionization detectors and catalytic sensors are commonly used. Other instruments, such as colorimetric instruments, may also be used as long as they can demonstrate that it can detect a vapor reading of 500 parts per million (ppm) for gasoline or 50 ppm for diesel vapors. However, each instrument has its own inherent limitations that must be overcome in the vapor monitoring process to avoid spurious readings. The VMP must include sufficient information to indicate how the sampling method will compensate for those limitations. Flame Ionization Detectors (FID):FIDs can exhibit flame-outs due to low oxygen levels within the well. Therefore, dilution of well vapors may be necessary to increase oxygen levels and avoid flame-outs. The instrument is also sensitive to methane, which may result from natural sources or the biodegradation of existing petroleum contamination. A carbon filter must be used to eliminate methane from the total hydrocarbon reading. Catalytic Sensors:Catalytic sensors are quenched by low oxygen levels making dilution necessary. As with FIDs, the presence of methane can give false readings, making the use of a carbon filter necessary. However, catalytic sensors are also sensitive to high humidity in the well, and may indicate the presence of hydrocarbons without justification. Photo Ionization Detectors (PID):PIDs have limitations similar to an FID, but with greater complexity. High soil gas humidity, high levels of methane and the alkane constituents may affect this instrument's response. Therefore, sampling should always be performed with and without dilution. In addition, PIDs show a limited response to diesel fuel vapors and a limited linear response to gasoline vapors. Consequently, the practical use of this instrument may be restricted.
(1) Calibration: Detectors shall be calibrated just prior to sampling the vapor monitoring wells at each facility. The manufacturer's instructions for calibration must be followed. A copy of the calibration procedure and a log containing the calibration date, gas used, the final instrument reading, and the initials of the technician performing the calibration shall be kept and maintained in accordance with the recordkeeping requirements of Rule 62761.710, F.A.C. (2) Sampling: The well with the lowest historical vapor reading should be sampled first. After each well is sampled, time must be allowed for the detector to purge itself of all vapors. A tight-fitting well plug, drilled and equipped with a tube to connect a pump and sampling instrument to the vapor monitoring well, must be used for each well sampling. The well plug should be inserted immediately after the removal of the well cap. When using detectors, the well should then be purged with a very low volume pump until the vapor level or oxygen concentration equalize. At least one and one half well casing volumes should first be purged before sampling. The well volume can be calculated be multiplying the depth of the well to groundwater by 3.14 times the radius of the well squared. Example: The vapor monitoring well is 10 feet deep (120 inches) and 2 inch diameter. The radius [r] is 1/2 the diameter, or 1 inch. well volume = 3.14 x r2 x depth well volume = 3.14 x 1 x 120 well volume = 376.8 cubic inches 565 cubic inches of air (1 and 1/2 volumes ) must be purged prior to sampling. To convert cubic inches to cubic feet divide the volume in cubic inches by 1728. In this example approximately 0.33 cubic feet of air would need to be purged. After the vapor monitoring well has been sufficiently purged, the detector or colorimetric instrument should be connected to the tubing and the sample drawn through it. The background reading should be taken at ground level just prior to connecting the detector to the hose on the monitoring well plug. The colorimetric instruments typically uses a tube which contains a material that changes color proportionately to the amount of a particular target compound present in the sample. Match the color of the tube to the closest on the
scale provided. Record the vapor readings on the Vapor Monitoring Sampling Record or equivalent. Dilution of the air flow from the well may be necessary to minimize spurious responses which could mask a release. Using an in-line "y" connector with the same diameter openings and equal length of tubing attached would dilute the air flow by a factor of two, resulting in a reading one half of the actual concentration. Higher levels of dilution can be obtained by using inexpensive flow meters (basically a plastic ball inside a calibrated tube) and plastic tubing clips. The dilution factor, as well as the undiluted and diluted readings should be recorded. Whenever present in the monitoring well, groundwater should be sampled at the groundwater/air interface and examined for turbidity and the presence of a layer of free product or a sheen. Any characteristics odors should also be noted. Observations should be documented on the Vapor Monitoring Sampling Record or equivalent. B. Release Detection Response Level (RDRL) Vapor readings will naturally fluctuate with environmental conditions. Changes in groundwater depth, subsurface humidity and temperature, for example, can cause increases in vapor readings even in the absence of a release. Given these factors, a statistical analysis should be performed to determine at what point vapor levels are indicative of a release. The RDRL is the statistical determination of the maximum vapor level expected for the site's normal conditions. Exceeding the RDRL would therefore suggest a release has occurred. Because site conditions vary from well to well, an RDRL shall be established for each well and/or probe. The following factors shall be considered in determining the RDRL: a. b. c. d. the the the the maximum monthly vapor concentration observed (Cmax) time-weighted average (Cave) standard deviation of the background vapor concentration readings (Sdev) time-related range in background vapor concentrations (Tr)
For diesel or gasoline or their equivalents, Cmax for any monitoring well or probe should not exceed 50 ppm for diesel or 500 ppm for gasoline. The limit for Cmax for any hazardous substance must be determined from the nature of the substance stored. Where Cmax exceeds the limit, further analysis must show that a release could be detected with statistical confidence. If Cmax is not within the 95% confidence interval, then that well or probe cannot be used for VM.
If Cmax is found to be within tolerance levels, then Cave and Sdev can be used to calculate time-related range in background vapor concentrations for each well or probe by the following formula: Tr = [Cave + 2(Sdev)] / [Cave - 2(Sdev)] Tr should be less than 5 for the well or probe to be suitable for vapor monitoring. In those cases where the ratio is equal to or greater than 5, further analysis must be used to show that a release could be detected with a statistical confidence interval of 95%. If Tr is not within the 95% confidence interval, then the well or probe cannot be used for vapor monitoring. If Cmax and Tr are within acceptable ranges, then the RDRL shall be calculated for each monitoring well or probe using the following formula: RDRL = Cave + 2(Sdev) D. Notification Procedure: A written protocol will be established for outlining the mechanism to be used in notifying the owners and operators as well as the Department or local program in the event the RDRL is reached or exceeded. The protocol will include names, position titles, and phone numbers. Any change in personnel must be submitted in writing; however, reapproval of the plan will not be necessary. III. DATA MANAGEMENT AND REPORTING PLAN
For the Site Suitability Assessment and the monthly monitoring phase, the data from each sampling event shall be collected at least every thirty days and shall include the following for each well (see APPENDIX I for sample record): a. Weather conditions (e.g., relative humidity and ambient air temperature) at the time of sampling/monitoring. b. Instrument used to perform the sampling. c. Depth to groundwater. d. Length of Screen above Water Table. e. Dilution Factor. f. Initial (undiluted/unfiltered) Reading. g. Filtered Reading. h. Diluted Reading. i. Final Reading. j. Background Vapor Reading k. Maximum product level in the tank since the last sampling event.
l. Name of person performing the sampling. m. Date of Sampling. The data collected in a. through m. above shall be evaluated for variations from the approved VMP. Statistical analyses, graphs and/or spreadsheets may be used in distinguishing suspected releases. The Department recommends that, at a minimum, the data obtained during the monthly monitoring phase be maintained in a spreadsheet format. However, if graphs are used, it is recommended that the following be shown as a function of time (i.e., month to month): a. Vapor concentration for each well, b. Ambient air and ground temperature, c. Ground water depth as a function of time, d. Ambient air relative humidity as a function of time, e. Any correlation between vapor reading and environmental parameters. If at any time the RDRL is exceeded, a Discharge Reporting Form (DRF) shall be submitted to the Department and the locally contracted program in accordance with Chapters 62761/762, F.A.C. The owner/operator must investigate in accordance with the federal, state, and local requirements. Within 7 working days following the submittal of the DRF, a Follow-up Incident Report (see APPENDIX II) must be completed and submitted to the Department. If the Follow-up Incident Report indicates that no release has occurred, the VMP must be amended to reflect the new data in order to avoid future false positive responses. ACKNOWLEDGMENTS This work was supported by the Florida Department of Environmental Protection, contract No. GC 348. We also wish to thank the members of the SOP Committee from the District and County office for the development and completion of this document. 1. Gary A. Robbins, PhD, "A Summary of Vapor Monitoring Issues Related to Release Detection Monitoring Using Total Organic Vapor Detection Instruments and Gas Chromatography in Florida and Recommendations for Resolving these Issues", (June 14, 1996).