Procedures for Sampling Hot Soil and Groundwater
Hot Soil Sampling Technique – Validity and Procedures
In 1993, a thermal remediation of a former gasoline service station was completed at the Lawrence
Livermore National Laboratory (LLNL). As part of this remediation, an innovative soil sample
analysis technique called Bulk Thermal Desorption (BTD) was used. BTD uses helium to purge the
entire soil core in a heated oven. The effluent vapors are then trapped and analyzed by conventional
means.
In the course of evaluating BTD, LLNL injected the hot soil samples with a spike of TCE and
chlorobenzene as the soil samples were collected. Although the principle intent of the spike was to
evaluate the performance of BTD, a secondary effect is that the spike allowed LLNL to also evaluate
the potential loss of VOCs during hot sample handling.
A total of 17 hot soil sample cores were spiked and the average spike recovery was 89% - this is
generally considered to be good recovery. LLNL also tested the spiking procedure on other cores and
concluded that the spike recovery “demonstrates that it just may not be possible to reliably spike a hot
soil sample in the field.” As a result of the spike tests, LLNL reported results for the hot soil sample
analyses without attempting to apply any correction factors due to potential VOC losses during
sample collection.
In 1999, Battelle Memorial Laboratories was hired by the US EPA, the Departments of Defense and
Energy, and the National Aeronautic and Space Administration to perform a re-evaluation of the
accuracy of hot soil sampling. Based on an average spike recovery of 94% Battelle concluded that
only a negligible loss of VOCs occurs during hot soil sampling.
Due to the LLNL and Battelle testing, subsequent thermal remediations have used hot sampling in
order to evaluate remedial effectiveness, using appropriate measures to reduce the potential for VOC
losses during sample handling.
Post remediation soil samples will most likely be collected using a hydraulic push rig due to the
shallow depth of sampling that is required. Hydraulic push sampling of hot soils is also inherently
safer than auger sampling because hydraulic push will not require operators to handle hot soil
cuttings. Although acetate liners are often used for hydraulic push core barrels at other sites, the
temperature of the subsurface will require the use of stainless steel, brass, or Teflon™ core barrel
liners.
Although the soil samples and core barrels will obviously be hot when extracted from the subsurface,
past thermal remediation experience has indicated that standard work gloves or heavy-duty rubber
gloves provide sufficient protection for the handling of hot push rods and soil core barrels.
Upon extracting the soil core barrel from the subsurface, the core barrel will be immediately capped
using the standard barrel caps provided by the hydraulic push manufacturer and the entire core barrel
will be placed on ice for cooling.
Figure 1. Soil core capped and cooling in an ice bath.
The thermometer is visible in the end cap.
When the core barrel is cool (typically 5-15 minutes), the core barrel is removed from ice and a soil
sub-sample is collected for analysis by the standard method. As in all soil sampling, care should be
taken to select the analysis sub-sample from near the center of the core barrel where evaporative
losses are minimized.
Figure 2. A soil sample being collected from along the
length of the core into a bottle containing methanol.
Hot Groundwater Sampling Technique
Low-flow sampling and purging techniques are used in an effort to collect the most representative
samples and to reduce the production of investigation-derived waste. Peristaltic pumps are used for
purging and sampling. A dedicated 1/4-inch Teflon™ sample tube is installed in each monitoring well
before remediation begins. The tubing sample inlet is set in the well at the center of the saturated
screen length of the well. Each sample tube has a 1/4-inch sample valve above the wellhead as shown
in Figure 3 below.
The wellhead completion (Fig. 3) forms a pressure tight boundary. The following low-flow
groundwater sampling procedure is adapted from the methods provided in the B&R Environmental,
Technical Memo for Purging and Groundwater Sampling Using Low Flow Purging and Sampling
Techniques (B&R Environmental, 1998). The apparatus used to perform sample cooling is shown in
Figure 4 below.
Figure 3. Wellhead Completion
Prior to initial sampling, a cooling coil is formed by wrapping a 10-ft length of ¼-inch stainless steel
tubing around a 4-inch diameter pipe until 6 full turns have been made. The ends of the tubing are
fashioned such that both ends of the tubing extend upward, as shown in Figure 4.
Figure 4. Sample Cooling Apparatus
- 1. Telephone the TRS operator the day prior to sampling to schedule a remote shutdown. A
shutdown period of at least 12 hours is preferred prior to groundwater sampling. - Lock-out and tag-out the ERH PCUs as described in a site-specific Health and Safety Plan. Note:
this procedure can only be completed by personnel who have been trained and certified by TRS in
lock-out and tag-out procedures. - Connect ¼-inch sample tubing to the cooling coil and place the coil in a bucket or cooler with ice
to form the ice bath. - Connect the cooling coil and peristaltic pump to the monitoring wellhead.
- Purge the well at an initial rate of 1 liter per minute until the field indicator parameters (see Step
6) stabilize and the minimum purge volume is removed. The minimum purge volume is two times
the static saturated well volume. The equation to calculate the minimum purge volume is:
V = 7.48κrw2(td-12)
Where V = one purge volume in gallons; rw = radius of well casing in feet; td = total depth of
well in feet; 12 = typical depth to groundwater in feet.
TRS Sampling Hot Media 031008 acf 5 - The pumping rate is recorded on purge data sheets every 3 to 5 minutes during purging. Any
adjustments to the pumping rate are recorded. Pumping rates should, as needed, be reduced to the
minimum capabilities of the pump (i.e., 0.1 to 0.2 liter per minute) to ensure stabilization of
parameters. Adjustments to the pumping rate are best made within the first 15 minutes of purging
to minimize purging time. - At the initiation of well purging and during the purging effort, water quality parameters including
turbidity, specific conductance, pH and dissolved oxygen (DO) are measured with a Horiba (or
equivalent) meter with flow cell. Readings are recorded on the purge data sheets every 3 to 5
minutes. Field parameters are monitored until stabilization occurs. Stabilization is complete when
three consecutive readings are within the following criteria:- Specific conductance and DO readings within 10 percent
- pH within +/-0.2 standards units
- Turbidity at 10 NTUs or less
- After the minimum purge volume is purged and all water quality parameters have stabilized,
sampling may begin. If all parameters have stabilized, but turbidity remains above 10 NTUs,
decrease the pump rate and continue monitoring. If the pump rate cannot be reduced and turbidity
remains above 10 NTUs, the information will be recorded and sampling begun. For low yield
wells, sampling commences as soon as the well has recovered sufficiently to collect the
appropriate volume for the anticipated samples. - VOCs are collected first utilizing the following method: a column of water is drawn in the
cooling coil tubing with the pump; the well sample valve and the peristaltic pump inlet valve are
closed and the pump shut off; the cooling coil is disconnected from the well sample valve; the
cooling coil is carefully removed from the ice bath; the pump inlet valve is opened; the sample is
decanted into the sample vials from the pump end of the tubing via gravity flow. The process is
repeated until the sample volume is collected. - Any other sample fractions (cations, anions) are sampled from the well end of the cooling coil
tubing. - Groundwater samples including quality control (QC) samples are labeled, preserved and shipped
per the Clients Sampling and Analysis Plan.
Groundwater samples are submitted to an outside laboratory for analysis. It is anticipated that one
groundwater sample per well plus the associated QC samples are collected and submitted to the
laboratory for analysis. A 7-day turnaround time (TAT) is requested for all analytical results.