Leaching of Chlorinated Phenols from Creosote NAPL-Impacted Soils and Soil–Cement Mix Designs

Abstract

This paper presents the results of a laboratory-based treatability study conducted for a confidential former wood treating site heavily impacted by a creosote non-aqueous-phase liquid (NAPL) containing pentachlorophenol (PCP). PCP impacts in the silty sands extended to approximately 33 ft (10 m) below the ground surface (bgs), with discrete soil samples containing PCP concentrations up to 14,500 mg/kg, and groundwater PCP concentrations forming a main plume exceeding 1 mg/L over 2.16 acres (0.87 ha). Treatability testing was performed on unspiked and NAPL-spiked site soils with total PCP concentrations ranging from 10 to 100 mg/kg, respectively, and leachable PCP concentrations of approximately 3 to 8 mg/L. Stabilization/solidification (S/S) mix designs using 5 to 10 weight percent (wt%, dry-reagent-to-wet-soil mass basis) of a Portland cement (PC) blend and 1 wt% powdered bentonite met the minimum unconfined compressive strength (UCS) and maximum hydraulic conductivity (K) performance criteria of 50 lb/in2 (345 kPa) and 1 × 10⁻⁶ cm/s, respectively, within the specified 28-day cure time. Long-term semi-dynamic leach testing was performed on S/S-treated soils using a modified United States Environmental Protection Agency (EPA) Method 1315 test incorporating a polydimethylsiloxane (PDMS) liner to improve the data reliability for hydrocarbons. Results showed that adding 1 wt% organoclay (OC) to the S/S mix designs did not substantially reduce leaching of common semi-volatile organic compounds (SVOCs) such as naphthalene, acenaphthene, phenanthrene and benzo(a)anthracene compared to mixes using only the PC blend with bentonite, consistent with previous studies. However, the inclusion of OC had a decisive effect on PCP immobilization, providing an order-of-magnitude (10×) reduction in the cumulative mass release of PCP over the test duration. This benefit diminished with decreasing degree of chlorination for other phenolic compounds.

1. Introduction and Background

This paper reports on the long-term leaching of semi-volatile organic compounds (SVOCs) from creosote non-aqueous-phase liquid (NAPL)-impacted soils from a confidential United States Superfund Site. Soils were treated by stabilization/solidification (S/S) and then tested by the United States Environmental Protection Agency (EPA) Method 1315 [1] modified (M) to enhance data reliability for use with organics. Key SVOCs included pentachlorophenol (PCP), naphthalene, and phenanthrene, as well as other polycyclic aromatic hydrocarbons (PAHs) with mass fractions greater than 1% in the NAPL.

PCP is frequently associated with wood treating sites [1,2,3,4] and its use often requires diesel as a carrier fluid, increasing its mobility when spilled/released. PCP itself has a solubility on the order of 14 mg/L up to its first acidity constant (pKa₁ = 4.59), and then its solubility increases in a log-linear manner to a maximum solubility of 243,000 mg/L around pH 9.1 [5]. For circumneutral soils near pH 7 to 7.5, the aqueous solubility of PCP varies between approximately 2000 to 6300 mg/L, which is two orders of magnitude greater than the key PAHs within the NAPL. Therefore, controlling PCP migration is a critical consideration for any effective remedial design.

Regardless of PCP, a common strategy in the remediation of creosote-NAPL-impacted soils is to consider the use of organophilic sorbents such as granular activated carbon (GAC) and organoclay (OC) products to sorb and immobilize the wide array of volatile organic compounds (VOCs) and SVOCs associated with creosote. Here, OC refers to a surface-modified high-swelling sodium bentonite clay impregnated with long-chain quaternary amines [6,7], transforming the clay into a hydrophobic medium capable of sorbing between 50 and 200 percent of its mass in NAPL [8].

Batch equilibration experiments such as shake extractions in deionized water [9] and the synthetic precipitation leaching procedure (SPLP) [10] have shown that PCP is strongly sorbed by OC after addition to a granular PCP-impacted soil, both without or with added cementitious reagents [6,7]. However, when tested as a monolith under long-term semi-dynamic conditions (EPA Method 1315M), the inclusion of OC (or GAC) in cement-based S/S treatments has not significantly reduced the leaching of VOCs, SVOCs and PAHs from manufactured gas plant (MGP) NAPL soils [11,12,13]. For example, Gentry et al. [13] showed that the inclusion of up to 4 percent by weight OC in S/S mixes only provided up to an additional 3% leaching reduction in the EPA Method 1315M cumulative mass release (CMR) of naphthalene over PC-only blends in MGP NAPL-impacted sediments. They concluded that the use of OC was not warranted because the small incremental reduction in CMR did not justify the significant reagent cost ($35/cubic yard; $45.6/m³ soil treated). EPA therefore dropped OC from consideration for the in situ S/S (or ISS) treatment of the Gowanus Canal sediments (Brooklyn, New York City, NY, USA), allowing millions of dollars to be re-directed to other aspects of the overall remedy. Similar arguments were made by Grubb and Briggs [14], who showed that the inclusion of up to 2 wt% GAC in S/S mix designs did not result in significant EPA Method 1315M CMR reductions in several pesticides (most notably lindane) to warrant the added cost of GAC.

Accordingly, since the suite of VOCs and SVOCs between MGP and creosote NAPL are very similar, it was fully expected by the authors that an S/S mix design value engineering (VE) study performed for the current project site would eliminate the proposed use of 1 wt% OC (based on SPLP testing) as part of the S/S treatment of 300,000 cubic yards (CY, or 229,500 m³) of soil. The estimated cost of 1 wt% OC across the entire ISS footprint was estimated at $13.37 million based on a reagent unit cost of $45/CY soil treated. Instead, the inclusion of OC had a decisive effect on PCP immobilization, but its impact diminished with decreasing degree of chlorination for other phenols such that no benefit was apparent for phenol. Thus, a very targeted deployment of OC was ultimately undertaken at the site, saving EPA an estimated $12.8 million in reagent costs.

2. Site and Testing Background

PCP was detected in site soils at concentrations up to 14,500 mg/kg. Figure 1 shows two (2003) groundwater plumes of PCP at the project site: a northern plume covering 2.16 acres (0.87 ha) with PCP exceeding 1 mg/L and a southern lobe covering 1.2 acres (0.49 ha) exceeding 5 mg/L. The proximity of the northern plume to the small creek on the northern property line was a major driver for site remediation.

Two treatability studies were performed to select S/S mix designs capable of effectively treating the creosote-NAPL-impacted soils at the project site. The original VE study performed by the authors in 2020 sought to eliminate OC from the recommended mix comprising 5 percent by weight (wt%) of a 30/70 weight-by-weight (w/w) blend of Type I/II Portland cement (PC) and a grade 100 ground granulated blast furnace slag cement (SC) with 1 wt% bentonite and 1 wt% OC. All doses are based on the dry weight of reagent to the wet weight of soil.

The VE study proposed by the authors instead used a 60/40 Type I/II PC/SC blend (which tends to develop strength more quickly) across four total S/S mix designs as summarized in

Table 1. Summary of mix designs tested between 2020 and 2023.

Short ID	Full ID	PC Type	PC/PLC	SC	Bentonite	OC	W/R
		--	wt%	wt%	wt%	wt%	L/kg
M1-2020	5.0PS(I-II)-OC (VE)	Type I/II	3.0	2.0	1.0	1.0	1.0
M2-2020	5.0PS(I-II) (VE)	Type I/II	3.0	2.0	1.0	--	1.0
M3-2020	7.5PS(I-II) (VE)	Type I/II	4.5	3.0	1.0	--	1.0
M4-2020	10PC(I-II) (VE)	Type I/II	6.0	4.0	1.0	--	0.9
M1	5.0PS(IL)-OC	Type IL	3.5	2.0	1.0	1.0	2.8
M2	7.5PS(IL)-OC	Type IL	4.5	3.0	1.0	1.0	2.8
M3	5.0PS(IL)	Type IL	3.0	2.0	1.0	--	2.8
M4	7.5PS(IL)	Type IL	4.5	3.0	1.0	--	2.8
M5	5.0PS(I-II)-OC	Type I/II	3.0	2.0	1.0	1.0	2.8
M6	5.0P(IL)-OC	Type IL	5.0	--	1.0	1.0	2.8
M7	7.5P(IL)-OC	Type IL	7.5	--	1.0	1.0	2.8
M8	5.0P(IL)	Type IL	5.0	--	1.0	--	2.8
M9	10PS(IL)-OC	Type IL	6.0	4.0	1.0	1.0	2.8
M10	10PS(I-II)-OC	Type I/II	6.0	4.0	1.0	1.0	2.8
I5A/I6A	I5A/I6A	Type I/II	3.0	2.0	1.0	--	2.8
I5A/I6A-OC	I5A/I6A-OC	Type I/II	3.0	2.0	1.0	1.0	2.8

NOTES: PC = Portland cement (Type I/II: MEDCEM 59 Houston Cement); PLC = Portland–limestone cement (Type IL: Holcim; Reserve, Louisiana Plant); SC = grade 100 slag cement (Holcim; Reserve, Louisiana Plant); PC/PLC and SC (where present) blended at a ratio of 60:40 on a dry-mass basis; Bentonite = sodium bentonite (Lonestar); OC = PM-199 organoclay (CETCO); W/R = final water-to-dry-reagent ratio at mixing; wt% = percent by weight; dry-reagent dose relative to wet-soil weight; L/kg = liters per kilogram; Mix ID = mix identification.

: 5, 7.5 and 10 wt% PC/SC with 1 wt% bentonite, and an additional 5 wt% PC/SC mix containing 1 wt% OC. The VE study reconfirmed that a minimum dose of 5 wt% of the 60/40 PC/SC blend with 1 wt% bentonite successfully passed the performance criteria for the site:

• Minimum unconfined compressive strength (UCS) of 50 pounds per square inch (lb/in²; 345 kPa) after 28 days of curing by ASTM D1633 [15] based on the average of triplicate samples.
• Maximum hydraulic conductivity (K) of 1.0 × 10⁻⁶ cm/s at 28 days curing by ASTM D5084 [16] based on single replicates.

A second treatability study was performed in 2023 shortly after the contractor mobilized to the site when a change in cement sources was required due to the limited availability of both Type I/II PC (which was being replaced by Type IL PC [PLC; limestone cement] in the US market) and SC. Also, the bucket mixing approach used by the contractor required that they operate at an elevated water-to-dry-reagent ratio (W/R) to successfully homogenize the soils in situ.

3. Materials and Methods

3.1. The 2020 Sampling Event

Fifteen gallons of bulk soil (three 5-gallon sealable steel buckets) were collected from borings drilled on 10-foot offsets in the vicinity of recovery well R-15 (see Figure 1), representing highly NAPL-contaminated soil from the ground surface to the top of the bedrock (approximately 30–32 ft or 9–10 m depth). An undisturbed soil sample was also collected at GP-6A (~19 ft or 5.8 m below ground surface [bgs]), outside of the contamination zone for measurement of field dry bulk density and K. Lighter-than-water NAPL (or LNAPL) was collected from recovery well R-13 and placed into 1 L glass jars (2) with Teflon-lined lids. Potable water from the Process Liquid Treatment System office building (Figure 1) was collected in 5.5 L LDPE cubitainers for use in S/S grout preparation. All sample buckets, jars, and cubitainers were stored in a walk-in refrigerator at approximately 6 degrees Celsius (°C) until testing.

3.2. The 2023 Sampling Event

Twenty gallons of bulk soil were collected in a similar manner as in the 2020 VE study near R-15. Since remediation construction was under way and the NAPL recovery well network was partially decommissioned, LNAPL was again collected from recovery well R-13, but also from R-21 and the east recovery trench cleanout) to ensure sufficient NAPL was recovered. A total of ten 1 L jars of a LNAPL/water mixture were collected; five jars from R-13, three jars from R-21, and two jars from the east recovery trench cleanouts. The NAPL was carefully decanted from each jar into a sealable 2 L glass jar yielding a total volume of approximately 1.65 L. Potable water was again collected as described above. All sample buckets, jars, and cubitainers were stored in a walk-in refrigerator at approximately 6 °C until testing.

Lastly, a freshly field-mixed ISS sample was collected during the early stages of ISS construction from cell “I5A/I6A” (see Figure 1) consisting of the 5 wt% PC blend dose with 1 wt% bentonite. Cylindrical molds of this sample were prepared both as-is and after adding 1 wt% OC to the mix ex situ to confirm field performance and evaluate the impact of OC on field-mixed samples.

3.3. Soil Sample Homogenization and Characterization

The 2020 soils were first homogenized by individual sample buckets by placing each chilled soil in a thick plastic bag and kneading and rolling the contents until visually homogeneous. The various individually homogenized media were then ultimately combined through alternating scoops of approximately equal sizes into a second series of bags that were again homogenized.

The 2023 soils were homogenized first by combining half of the site soil sample (approximately 10 gallons) in a large, galvanized steel tub and mixing with a mortar mixer. The mortar mixer was first used in an up-and-down motion to break up soil clods and homogenize the vertical profile, and then it was tilted and used in a side-to-side motion under the soil surface to incorporate the material together while limiting the amount of surface disturbance and volatilization until the final two passes incorporated the surface material. The remaining soil was homogenized in the same manner in a second galvanized tub. The homogenized soil from each tub was then evenly distributed between five 5-gallon buckets by alternating scoops of soil into each of the 5 buckets simultaneously, and the contents of each bucket were then homogenized in the steel tub as previously described before being returned to the bucket for storage.

After homogenization, representative subsamples of the bulk soil composites were collected and submitted for baseline geotechnical and as-is environmental characterization.

3.4. Soil Spiking with NAPL

Prior to leaching characterization and treatability testing, the homogenized soils were spiked with NAPL to increase their PCP concentration to a targeted 26 mg PCP/kg dry soil to match the 2015 study levels. The 2020 NAPL was dosed at 12 mL NAPL/kg dry soil. As the PCP concentration in the as-is soil in 2023 was comparable to the 2020 VE study and the NAPL PCP concentration was only slightly lower, the same spiking dose was used in the 2023 study. Incorporation of NAPL into the soil was performed in the same manner as the homogenization described for each of the studies (bag kneading in 2020 and mixing with a mortar mixer within a steel container in 2023); however, the batches were smaller (6 kg and 11 kg, respectively) to promote even NAPL distribution throughout the sample.

Subsamples were collected from the NAPL-spiked soils just after spiking for baseline environmental characterization. Untreated leaching tests and mix designs were performed after the soils equilibrated for 3 days.

4. Epa Method 1316MTesting

EPA Method 1316 [1] modified (M) for use with organics was used to evaluate the liquid–solid partitioning of the soil surrogates in a series of batch equilibrium tests at different liquid-to-solid ratios (L/S). Consistent with Khuri et al. [17], two main modifications were used during sample preparation and processing to increase data reliability for hydrocarbon analysis: (1) samples were not dried prior to testing and the MC was factored into the L/S at setup and (2) extractions were performed under zero-headspace conditions.

In 2020, the extractors were tumbled at approximately 30 revolutions per minute for 24 h plus or minus (±) 2 h, in accordance with method specifications based on the sample grain size. The contact time was extended to 7 days in the 2023 study to promote equilibration with the NAPL. Eluate pH (EPA Method 9040C [18]) and conductivity (EPA Method 9050 [19]) were immediately measured after the extractors were opened for processing. The waters were then filtered through a 0.7 µm glass-fiber Toxicity Characteristic Leaching Procedure (EPA Method 1311) syringe filter for VOC (EPA Method 8260D [20]) and SVOC (EPA Method 8270E [21]) analysis to remove any particulates or suspended NAPL, and the dissolved organic carbon (DOC) sample was further filtered through a 0.45 µm syringe filter before analysis by Standard Methods (SMs) 5310B [22].

A replicate of L/S 20 was added since it is often difficult to obtain sufficient sample volume at low L/S ratios for SVOC testing. For context, an L/S of 0.25 to 0.35 is indicative of saturated soils with total porosities on the order of 30 to 40%. Only data from the 2023 soil samples will be reported here.

4.1. Laboratory Mix Designs

The 2020 and 2023 S/S mix designs are summarized by PC type, bentonite (B) and OC content along with the final W/R in Table 1. The reagent sources are identified in the notes of Table 1 recognizing that the PC blend was always a 60/40 PC/SC blend regardless of the actual use of PC (Type I/II) or PLC (Type IL). In all cases, the bentonite was hydrated overnight in the site potable water before the other reagents were added to make the grout immediately prior to soil blending. This was not the case for the field-mixed sample where the PC blend and bentonite were added at the same time in the grout plant. The OC was added in bulk to match the cell characteristics.

During mixing, some sheens were visible when the grout was initially added to the soil, but no free NAPL was evident upon blending. When the target texture was achieved (thick milkshake), subsamples of the mixture were collected for measurement of moisture content and organics (SVOCs), and then the remaining material was packed in 2 × 4-inch (5 × 10 cm) cylindrical molds. The cylindrical molds were sealed and allowed to cure at approximately 20 °C in a moisture-controlled box until the molds were submitted for testing as described in the following sections.

4.2. Geotechnical Tests

UCS was measured in triplicate at 7 and 28 days of curing for all mixes and in duplicate at 56 days of curing for select 2023 laboratory mixes (M1–M8) using test cylinders that were 2 × 4 inches (5 × 10 cm) in size. Testing was performed in accordance with ASTM D1633.

K testing was performed on 28-day-cured test cylinders that were 2 × 4 inches (5 × 10 cm) in size. Tests were performed according to ASTM D5084 using a flexible wall permeameter with a confining stress of 8.7 lb/in² (60 kPa) to simulate the stress at the mid-height of the ISS mass in situ.

4.3. EPA Method 1315M Semi-Dynamic Leach Testing

EPA Method 1315M semi-dynamic leach testing was performed on select 28-day-cured mix designs to measure the SVOC mass released from the cylindrical molds in the radial (3D) configuration. Modifications to EPA Method 1315 (similar to NEN 7375 [23]) were used to increase data reliability for organic compounds owing to their volatility and/or limited aqueous solubility in water. For the unmodified leaching test, the sample material is suspended in a bath of deionized (DI) water at a liquid-to-surface-area ratio (L/Sa) of 9 ± 1 mL/cm² for a specified duration, and then it is removed and placed in a fresh DI water bath at the same L/Sa. The test uses nine time intervals (denoted T1 to T9), with exchanges occurring after 2 h and after 1, 2, 7, 14, 28, 42, 49, and 63 cumulative days of leaching. At the end of each leaching interval, the eluate water is subsampled for analysis.

The main modifications for organics testing include maintenance of zero-headspace conditions and the use of a polydimethylsiloxane (PDMS) jar liner [24] to sorb organic compounds, maintaining a strong gradient for mass transfer out of the solids and into the surrounding water. The deployment of the PDMS liner means the total mass leached from an S/S-treated sample during each interval is the sum of the compound mass in the water eluate and the PDMS (i.e., water + PDMS), and therefore the mass sorbed to the PDMS must be determined. To do this, the eluate water is sampled for the analytical suite and then discarded, and the PDMS liner is extracted over a 16 h period of longitudinal rolling with 30 mL of either methanol (for VOC analysis) or acetonitrile (for SVOC analysis). At T6 (28 days of testing, after a 2-week leaching interval), the extraction efficiency factor (EEF) of each compound from the PDMS is determined through four serial extractions of the PDMS for each mix design sample. In general accordance with Gentry et al. [13] and Khuri et al. [17], an EEF can be calculated for any analyte with two or more detections in the four-extract series by dividing the mass removed in the first extract by the sum of the mass removed in four extracts. The average EEF calculated for each individual compound was applied as a correction factor to the single extraction mass measured at each exchange event. The corrected mass from the PDMS was then combined with the corresponding mass from the water bath (thus, water + PDMS) to obtain the total mass released.

The DI water baths were analyzed for pH (EPA 9040C), conductivity (EPA 9050A), SVOCs (EPA 8270E-SCAN and -SIM) and dissolved organic carbon (SM 5310B) for all nine exchange events, while the acetonitrile extract of the PDMS liner was submitted for SVOC analysis by EPA 8270E-SCAN/SIM. VOCs were also analyzed for in the water and methanol matrices by EPA 8260D in the 2020 study only.

5. Results and Discussion

5.1. Geotechnical Baseline Characterization

Baseline geotechnical results for the untreated soil are summarized in

Table 2. Geotechnical characterization of bulk homogenized soils.

Soil ID	MC	LOI		Grain Size Distribution					Atterberg Limits
	ASTM D2166	ASTM D2974		ASTM D6913/7928/2487					ASTM D4318
	MC (wt%)	MC (wt%)	LOI (wt%)	Sample Description	Classification (USCS)	Gravel (wt%)	Sand (wt%)	Fines (wt%)	LL	PL	PI
2020 Bulk Sample	N/M	20	1.4	Silty Sand	SM	2.1	64.4	33.5	Non-Plastic
2023 Soil Replicate 1	24.1	15	2.2	Silty Sand	SM	0.7	67.4	31.9	Non-Plastic
2023 Soil Replicate 2	22.3	16	1.6	Silty Sand	SM	2.6	63.8	33.6	Non-Plastic

NOTES: ASTM = ASTM International; PL = Plastic Limit. LL = Liquid Limit. SM = silty sand. LOI = Loss on Ignition. USCS = unified soil classification system; MC = moisture content; wt% = weight percent (oven-dried-soil basis); PI = Plasticity Index.

. The 2020 soil sample was characterized as a non-plastic silty sand (SM soil). The undisturbed field sample from GP-6A had a dry bulk density (DBD or unit weight) of 101.9 lb/ft³ (16 KN/m³) and a hydraulic conductivity of 9.33 × 10⁻⁵ cm/s. The 2023 soil sample taken from the same general area and depths was essentially the same.

5.2. Environmental Baseline Characterization

Baseline environmental results for the NAPL, as-is and NAPL-spiked soils, and potable water are respectively shown in

Table 3. Non-aqueous-phase liquid characterization (detects and phenolics only).

	Analyte	Units	NAPL (2020)		NAPL (2023)
	General Chemistry
	Density	g/mL	0.968		0.976
	Volatile Organic Compounds by SW8260D
	Xylene (total)	mg/kg	3.78		0.022	U
	Semi-volatile Organic Compounds by SW8270E-SCAN
Phenolics	Pentachlorophenol	mg/kg	14,600		10,600
	2,3,4,6-Tetrachlorophenol	mg/kg	718		590
	2,4,5-Trichlorophenol	mg/kg	126	J	88.7
	2,4-Dimethylphenol	mg/kg	89.0	U	24.0	U
	2,4-Dichlorophenol	mg/kg	0.2	U	13.0	U
	3&4-Methylphenol	mg/kg	41.0	U	9.6	J
	2,4,6-Trichlorophenol	mg/kg	0.1	U	9.2	U
	2-Methylphenol	mg/kg	32.0	U	8.9	U
	2-Chlorophenol	mg/kg	0.1	U	8.2	U
	Phenol	mg/kg	26.0	U	3.9	U
Other SVOCs	Phenanthrene	mg/kg	40,800		34,800
	Fluoranthene	mg/kg	16,200		17,300
	Fluorene	mg/kg	12,300		8980
	Acenaphthene	mg/kg	11,900		9110
	Naphthalene	mg/kg	11,300		6110
	Pyrene	mg/kg	9800		7660
	Dibenzofuran	mg/kg	9030		6380
	Anthracene	mg/kg	3910		3600
	2-Methylnaphthalene	mg/kg	3270		2300
	1-Methylnaphthalene	mg/kg	2590		2920
	Chrysene	mg/kg	2220		2140
	Benzo(a)anthracene	mg/kg	1950		2350
	1,1′-Biphenyl	mg/kg	1780		1690
	Benzo(b)fluoranthene	mg/kg	1060		1090
	Carbazole	mg/kg	659		742
	Benzo(a)pyrene	mg/kg	551		459
	Benzo(k)fluoranthene	mg/kg	393		357
	Acenaphthylene	mg/kg	293		220
	Indeno(1,2,3-cd)pyrene	mg/kg	202		146
	Benzo(g,h,i)perylene	mg/kg	148		140
	bis(2-Ethylhexyl)phthalate	mg/kg	60.2	J	51.8
	Dibenz(a,h)anthracene	mg/kg	59.2		66.4

NOTES: NAPL = non-aqueous-phase liquid. After phenolics, semivolatile organic compounds are ordered from greatest to least mass as measured in the NAPL. g/mL = grams per milliliter. mg/kg = milligrams per kilogram. B = Compound was found in the blank and sample. U = The analyte was analyzed for but not detected. J = Result is less than the reporting limit but greater than or equal to the method detection limit and the concentration is an approximate value. Italicized values denote EPA 8270-SIM.

,

Table 4. Baseline characterization of as-is and NAPL-spiked surrogate soils (project detects only).

			2015 Data		As-Is Soil				NAPL-Spiked Soil
					2020		2023		2020		2023
	General Chemistry
	NAPL Addition	mL/kg			--		--		12		12
	pH (SW9045D)	SU	NT		NT		NT		NT		6.97
	Solids Content	wt%, g/g-wet	NT		82.3		81.2		82.0		81.5
	Moisture Content	wt%, g/g-dry	NT		21.5		23.3		22.0		22.7
	Volatile Organic Compounds by EPA SW8260D
	1,2,3-Trichlorobenzene	mg/kg	0.354	U	0.18	U	9.70	U	1.80	U	9.20	U
	1,2,4-Trichlorobenzene	mg/kg	0.354	U	0.18	U	9.70	U	1.80	U	9.20	U
	2-Butanone (MEK)	mg/kg	4.72	U	0.27	U	9.45	U	2.70	U	8.90	U
	Acetone	mg/kg	2.36	U	0.29	U	16.00	U	2.90	U	15.50	U
	Benzene	mg/kg	0.354	U	0.104		1.800	U	0.330	U	1.650	U
	Carbon disulfide	mg/kg	NR		0.067	U	2.100	U	0.680	U	1.950	U
	Chlorobenzene	mg/kg	0.354	U	0.111	J	1.800	U	0.335	U	1.650	U
	Ethylbenzene	mg/kg	1.09	J	1.24		1.80	U	1.17		2.28	U
	Isopropylbenzene	mg/kg	0.354	U	0.578		5.500	U	0.527	J	5.200	U
	m,p-Xylene	mg/kg	3.57		4.35		4.96		3.84		6.75
	Methyl acetate	mg/kg	NR		0.1	U	5.40	U	1.00	U	5.10	U
	Methylene chloride	mg/kg	0.354	U	0.18	U	10.00	U	1.80	U	9.60	U
	o-Xylene	mg/kg	1.6		2.1		2.02	J	1.78		2.49	J
	Styrene	mg/kg	0.564	J	1.01		1.600	U	0.799	J	1.450
	Toluene	mg/kg	1.19		0.84		2.000	U	0.874		1.950	U
	Xylene (total)	mg/kg	5.17		6.45		6.98		5.62		8.38
	Semi-volatile Organic Compounds by EPA SW8270E-SCAN
Phenolics	Pentachlorophenol	mg/kg	25.6		8.6		4.5		54.55		105.05
	2,3,4,6-Tetrachlorophenol	mg/kg	NR		0.944	J	1.2		2.965	J	4.025
	2,4,5-Trichlorophenol	mg/kg	0.782	U	0.15	U	0.8	U	0.575	U	0.15	U
	2,4,6-Trichlorophenol	mg/kg			0.12	U	0.7	U	0.46	U	0.12	U
	2,4-Dichlorophenol	mg/kg			0.17	U	0.9	U	0.655	U	0.17	U
	2,4-Dimethylphenol	mg/kg	7.30		1.44		6.6		1.4	U	4.78
	2-Chlorophenol	mg/kg			0.099	U	0.5	U	0.38	U	0.099	U
	2-Methylphenol	mg/kg	2.19		1.59		2.8		0.5115	J	1.805
	3&4-Methylphenol	mg/kg	3.15		2.25		4.2		1.0775	J	2.62
	Phenol	mg/kg	0.782	U	1.02		1.9		0.453		1.3405
Other SVOCs	Phenanthrene	mg/kg	573		670		469.5	E	579.5		1036
	Fluoranthene	mg/kg	318		262		285.0		226		325.5
	Fluorene	mg/kg	215		195		199.5		180		236
	Acenaphthene	mg/kg	191		174		198.5		177		233.5
	Naphthalene	mg/kg	578		634		1092.0	E	445		947
	Pyrene	mg/kg	188		141		174.5		149		201
	Dibenzofuran	mg/kg	153		147		140.5		132.5		172.5
	Anthracene	mg/kg	167		95.8		115.4		71.45		94.7
	2-Methylnaphthalene	mg/kg	175		146		189.0		119.5		153.7
	1-Methylnaphthalene	mg/kg	NR		62.5		102.5		51.25		80.8
	Chrysene	mg/kg	47.2		43.8		50.1		34.2		50.45
	Benzo(a)anthracene	mg/kg	51.3		41.7		55.5		34.55		51.2
	1,1′-Biphenyl	mg/kg	NR		35.4		38.6		32.8		41.35
	Benzo(b)fluoranthene	mg/kg	38.8		17.6		34.5		18.4		31.6
	Carbazole	mg/kg	93		60.8		54.4		35.7		45.65
	Benzo(a)pyrene	mg/kg	25.0		11.4		24.3		11.75		22.65
	Benzo(k)fluoranthene	mg/kg	12.5		5.98		13.9		6.175		8.6
	Acenaphthylene	mg/kg	8.41		5.37		11.3		5.82		8.48
	Indeno(1,2,3-cd)pyrene	mg/kg	8.94		4.31		9.9		3.055		3.055
	Benzo(g,h,i)perylene	mg/kg	8.8		3.51		10.4		3.235		3.235
	bis(2-Ethylhexyl)phthalate	mg/kg	0.782	U	0.047	U	0.3	U	0.18	U	0.175	J
	Dibenz(a,h)anthracene	mg/kg	2.7		0.971		3.21		0.961		2.63
	Hexachlorobenzene	mg/kg			0.051	U	0.28	U	0.19	U	0.051	U
	Aniline	mg/kg	0.782	U	NR		1.0		0.4785	J	0.8245
	Pyridine	mg/kg	0.782	U	NR		0.4	J	0.26	U	0.094	J

NOTES: Reported data is for the average of two soil replicates except the 2020 as-received soil. After phenolics, semivolatile organic compounds are ordered from greatest to least mass percentage as measured in the NAPL (see Table 2). °C = degrees Celsius. SU = standard pH units. NT = Not Tested. NR = Not Reported. E = Exceeds calibration range. B, J, U = See Table 3 notes. Italicized values denote by EPA 8270-SIM.

and

Table 5. Water mix characterization.

Parameter	Method	Units	2020	2023
General Chemistry
pH	SW9040	SU	7.59	8.33
Alkalinity, Bicarbonate	SM4500D	mg/L	256	262
Alkalinity, Total as CaCO3	SM2320B	mg/L	257	266
Sulfate	EPA 300.0	mg/L	6.8	3.7
Chloride	EPA 300.0	mg/L	60.5	67.1

NOTES: Mix water is tap water from the Process Liquid Treatment System (PLTS) building. SU = standard pH units. mg/L = milligrams per liter. CaCO₃ = calcium carbonate.

. The compositional results for the creosote NAPL are shown in Table 3 for the VOC/SVOC analytes detected above the method detection limit only, except for numerous phenols since they were of specific interest in this study. The NAPL samples contained no detectable VOCs other than xylene. The SVOCs are organized first by phenolics in order of decreasing degree of chlorination, and then by mass fraction in the NAPL. The 2020 NAPL had phenolics and other key SVOCs that were about 50% higher than the 2023 NAPL sample. The compositional difference between the two NAPLs likely reflects that the 2023 NAPL was collected from multiple site sources. The SVOCs shown in Table 3 and their respective chemical properties were used to estimate their equivalent mass and mole fractions in the NAPL, as shown in Table S1, to differentiate between the key compounds (e.g., greater than 1 wt% in NAPL) and trace compounds. Most key SVOCs (Table S1) were relatively similar between the two NAPL sources (Table 3).

Table 4 shows the general chemistry, VOC and SVOC results for the homogenized as-is and NAPL-spiked soil surrogates where all values are the average results of duplicate samples except the 2020 as-is soil (singlet). The prior 2015 study results and respective 2020/2023 NAPL dosing rates are shown for reference. The PCP levels of the NAPL-spiked soils were considerably higher than the targeted 26 mg/kg, in part to account for potential volatilization losses during material handling and testing. Whereas several lesser chlorinated and methylated phenols were not evident in the discrete NAPL samples (Table 3), Table 4 shows that they did occur in the surrogate soils, and were therefore expected to appear in the leachates from the mix designs unless either immobilized or reacted (e.g., transformed by alkaline hydrolysis similar to lindane [14]). Note that trichlorophenols were not detected in the surrogate soils despite being added via the NAPL (Table 3) used for spiking activities.

The potable site water for the S/S mix designs had a pH of 7.6 to 8.3, alkalinity between 250 and 270 mg/L as calcium carbonate and low levels of chlorides and sulfate (Table 5).

6. Epa Method 1316M Results

The eluate analytical results are summarized by L/S replicate in

Table 6. EPA SW1316M results for 2023 NAPL-spiked surrogate soil.

	Parameter/Analyte	Method	Units											Select Effective Solubilities, Ceff
	Liquid/Solid Ratio		L/kg	20		10		5		2		1
	General Chemistry
	pH	SM4500	SU	7.8		7.54		7.34		7.04		6.95
	Conductivity	SM2510B	mS/cm	0.80		1.43		2.05		2.66		2.76
	Dissolved Organic Carbon	SM5310	mg/L	18.3		24.9		39.9		74.3		106
	VOCs
	Styrene	SW8260D	mg/L	0.014		0.0186	J	0.0212		0.0245		0.0214		0.020
	Xylene (total)	SW8260D	mg/L	0.0943		0.111		0.119		0.126		0.114		0.11
	m,p-Xylene	SW8260D	mg/L	0.0611		0.072		0.0756		0.0809		0.0741		0.073
	o-Xylene	SW8260D	mg/L	0.0332		0.0385		0.0429		0.0448		0.0403		0.040
	Toluene	SW8260D	mg/L	0.0202		0.0287		0.034		0.0465		0.0394		0.034
	Ethylbenzene	SW8260D	mg/L	0.0199		0.0228		0.0233		0.028		0.0243		0.024
	SVOCs
Phenolics	Pentachlorophenol	SW8270SIM	mg/L	3.8		6.01		6.69		8.14		4.62		6.16
	2,3,4,6-Tetrachlorophenol	SW8270E	mg/L	0.181	J	0.342	E	0.762		0.621		0.32
	2,4,5-Trichlorophenol	SW8270E	mg/L	0.0295		0.0291		0.0337		0.0299		0.0157
	2,4,6-Trichlorophenol	SW8270E	mg/L	0.0014	J	0.0018	J	0.0029	J	0.0043	J	0.0025	J
	2,4-Dichlorophenol	SW8270E	mg/L	0.0037	J	0.0042		0.0066		0.001	U	0.0058
	2,4-Dimethylphenol	SW8270E	mg/L	0.112		0.155		0.46		0.698		0.571
	2-Chlorophenol	SW8270E	mg/L	0.00074	U	0.00074	U	0.0012	J	0.00074	U	0.00074	U
	2-Methylphenol	SW8270E	mg/L	0.0769		0.113		0.345		0.629		0.593
	3&4-Methylphenol	SW8270E	mg/L	0.0661		0.101		0.216		0.582		0.548
	Phenol	SW8270E	mg/L	0.0059		0.0091		0.0176		0.0409		0.0301
Other SVOCs	Phenanthrene	SW8270SIM	mg/L	0.189		0.146		0.092		0.102		0.0755		0.12
	Fluoranthene	SW8270SIM	mg/L	0.0212		0.0172		0.00867		0.00952		0.006		0.013
	Fluorene	SW8270SIM	mg/L	0.183		0.149		0.129		0.143		0.0977		0.14
	Acenaphthene	SW8270SIM	mg/L	0.312		0.3		0.209		0.243		0.165		0.25
	Naphthalene	SW8270E	mg/L	2.63		2.3		2.43		3.22		1.93		2.32
	Pyrene	SW8270SIM	mg/L	0.0175		0.0102		0.00508		0.00508		0.00282
	Dibenzofuran	SW8270E	mg/L	0.17	J	0.153		0.128		0.129		0.0769		0.13
	Anthracene	SW8270SIM	mg/L	0.0228		0.0188		0.0149		0.0184		0.0101		0.02
	2-Methylnaphthalene	SW8270SIM	mg/L	0.369		0.35		0.253		0.291		0.213		0.30
	1-Methylnaphthalene	SW8270SIM	mg/L	0.244		0.19		0.168		0.192		0.13		0.18
	Chrysene	SW8270SIM	mg/L	0.00123		0.00088		0.00029	J	0.0006		0.00027	J	0.00065
	Benzo(a)anthracene	SW8270SIM	mg/L	0.00157		0.00101		0.00047		0.00084		0.00037		0.00085
	1,1′-Biphenyl	SW8270E	mg/L	0.0737		0.0514		0.0456		0.0514		0.0313		0.051
	Benzo(b)fluoranthene	SW8270SIM	mg/L	0.00087		0.00039		0.00012	U	0.00023		0.00012	U
	Carbazole	SW8270E	mg/L	0.329		0.229		0.485		0.348		0.203		0.319
	Benzo(a)pyrene	SW8270SIM	mg/L	0.00048		0.00018	J	0.00012	U	0.00014	J	0.00012	U
	Benzo(k)fluoranthene	SW8270SIM	mg/L	0.00025	J	0.0002	U	0.0002	U	0.0002	U	0.0002	U
	Acenaphthylene	SW8270SIM	mg/L	0.0155		0.0152		0.0136		0.0159		0.0115		0.0143
	bis(2-Chloroethyl)ether	SW8270E	mg/L	0.0011	J	0.00086	U	0.00086	U	0.00086	U	0.00086	U
	Pyridine	SW8270E	mg/L	0.0027	J	0.0028	J	0.0012	J	0.00062	U	0.0023	J

NOTES: After phenolics, SVOCs are ordered from greatest to least mass percentage as measured in the NAPL (see Table 3). SU = standard pH units. J, U, E = See Table 3 and Table 4 notes.

for the 2023 surrogate soils only for purposes of brevity. Eluate pH ranged between 6.9 and 7.8 standard units and SVOC concentrations typically increased with decreasing L/S. DOC concentrations ranged from 18 to 106 mg/L. PCP concentrations ranged from 3.8 mg/L at an L/S of 20 to a maximum of 8.14 mg/L at an L/S of 2, comparable to the PCP concentrations in the southern lobe and exceeding the central PCP plume (Figure 1).

Figure 2 shows several SVOC data using format wherein the total mass release (concentration x L/S) is plotted against the L/S of the corresponding replicates. In this format, the mass release of a compound at any given L/S is bound by its respective total content (from Table 4) and the plotted curves are described as solubility-controlled for those portions of its curve that are characterized by a 1:1 slope. Invariant or flat slopes are thus available-content controlled per EPA Method 1316. In Figure 2A, the total PCP concentration is 105 mg/kg (from Table 4) which defines the maximum available-content controlled conditions and its release curve is characterized by an effective concentration (C_eff) of 5.85 mg/L, based on a linearization of its L/S 1 to 20 data for the 2023 treatability sample. Such C_eff data are needed to perform solubility checks on the EPA Method 1315M water baths to ensure that strong gradients for mass transfer persist through the long-term leaching tests. The corresponding EPA Method 1316M mass release data and effective solubility determinations for other key SVOCs such as naphthalene, phenanthrene and benzo(a)anthracene are shown in Figure 2B–D. These SVOCs were chosen for presentation purposes due to their common occurrence at creosote sites, but more importantly because the range of octanol–water partitioning coefficients (K_ow) for these SVOCs (3.7 to 5.5) overlaps with that of PCP (4.74), implying that the greater hydrophobicity of benzo(a)anthracene over PCP should correspond to greater affinity for OC. Additional C_eff data are reported in Table 6 for those key SVOCs where 1:1 linearization of their 2023 treatability sample mass release data was possible.

7. Mix Design Results

7.1. Hydraulic Conductivity

As summarized in

Table 7. Summary of ISS hydraulic conductivity (K) and unconfined compressive strength (UCS) for 28- or 56-day-cured samples.

ID	PC/PLC Type	PC or PLC	SC	B	OC	W/R	DBD		28-Day K	28-day UCS		56-Day UCS
		wt%	wt%	wt%	wt%	L/kg	kN/m3	lb/ft3	cm/s	kPa	lb/in2	kPa	lb/in2
M1-2020	Type I/II	3.0	2.0	1.0	1.0	1.0	14.8	94.3	1.9 × 10−7	1758 ± 572	255 ± 83	NT
M2-2020	Type I/II	3.0	2.0	1.0	--	1.0	15.4	98.0	2.2 × 10−7	3089 ± 731	448 ± 106
M3-2020	Type I/II	4.5	3.0	1.0	--	1.0	14.6	93.0	1.7 × 10−7	4606 ± 83	668 ± 12
M4-2020	Type I/II	6.0	4.0	1.0	--	0.9	14.5	92.5	2.2 × 10−7	2358 ± 1393	342 ± 202
M1	Type IL	3.0	2.0	1.0	1.0	2.8	11.7	74.4	1.1 × 10−7	545 ± 34	79 ± 5	614 ± 127	89 ± 18.4
M3	Type IL	3.0	2.0	1.0	--	2.8	12.0	76.7	1.5 × 10−7	552 ± 34	80 ± 5	710 ± 54	103 ± 7.78
M2	Type IL	4.5	3.0	1.0	1.0	2.8	10.9	69.6	5.0 × 10−7	414 ± 34	60 ± 5	517 ± 20	75 ± 2.83
M4	Type IL	4.5	3.0	1.0	--	2.8	11.2	71.3	1.6 × 10−7	476 ± 34	69 ± 5	676 ± 20	98 ± 2.83
M9	Type IL	6.0	4.0	1.0	1.0	2.8	NT			296 ± 131	43 ± 19	NT
M5	Type I/II	3.0	2.0	1.0	1.0	2.8	11.4	72.5	3.4 × 10−7	331 ± 62	48 ± 9	396 ± 24	57.5 ± 3.54
M10	Type I/II	6.0	4.0	1.0	1.0	2.8	NT			262 ± 7	38 ± 1	NT
M6	Type IL	5.0	--	1.0	1.0	2.8	11.1	70.5	5.7 × 10−7	179 ± 7	26 ± 1	203 ± 5	29.5 ± 0.71
M8	Type IL	5.0	--	1.0	--	2.8	11.4	72.4	5.3 × 10−7	172 ± 34	25 ± 5	265 ± 63	38.5 ± 9.19
M7	Type IL	7.5	--	1.0	1.0	2.8	10.6	67.3	7.2 × 10−7	186 ± 7	27 ± 1	238 ± 5	34.5 ± 0.71

NOTES: See Table 1 for reagent information. K = Permeability measured at 20 °C and effective stress of 8.7 lb/in². Reagent doses reported on a weight percent (wt%) basis with respect to wet-soil weight. UCS = unconfined compressive strength. W/R = final water-to-dry-reagent ratio at mixing. Average and standard deviation of UCS shown for triplicate samples. DBD = dry bulk density (reported as the average of triplicates).

, the K values ranged from 1.1 × 10⁻⁷ to 7.2 × 10⁻⁷ cm/s. All mixes met the site criteria of K ≤ 1 × 10⁻⁶ cm/s.

7.2. Unconfined Compressive Strength

The UCSs of the 28- and 56-day-cured ISS mix designs are summarized in Table 7 with a complete reporting of replicates, including 7-day UCSs and the sample dry bulk densities (DBDs) and moisture contents appearing in Table S2. The UCSs of the mix designs were very sensitive to the W/R ratio, with the W/R 2.8 mixes producing approximately one-fifth of the strength of the mixes with the same cement dose and W/R of approximately 1.0 at 28 days. Increasing the cement dose (Mixes M9 and M10) did not lead to higher strengths, likely due to their inclusion of even more water (higher cement dose adds more water per W/R). Portland cement type and use of OC had nominal impact on UCS; however, the inclusion of SC in mixes produced UCS values approximately 3 times higher than comparable mixes without SC at 28 days. Mixes 6 to 8 without SC never achieved the UCS criteria.

7.3. Epa Method 1315M

The EPA Method 1315M testing was initiated on 28-day-cured test cylinders and was performed identically for SVOCs in the 2020 and 2023 studies. Figure 3A shows the EPA Method 1315M pH measurements for all mix designs for all nine intervals for the 5 wt% PC blend (all), as well as the 7.5 and 10 wt% PC blends (2023 mixes only). The pH data for the 2020 VE study mixes at a W/R of 0.9 to 1.0, plotted using square symbols, are shown to be in the 10-to-11 range, whereas those with W/R of 2.8 from the 2023 study are typically 1 pH unit higher. The highest pH values were measured for PLC-only mixes. The corresponding EPA Method 1315M CMR curves for the same mixes are shown for naphthalene in Figure 3B. The data is tightly clustered with slightly higher CMRs being associated with the PLC. The 2020 VE study data, plotted using black square symbols, shows less naphthalene leaching over time compared to the rest of the mixes. This difference could be explained by the initial naphthalene concentration in the 2020 samples being approximately half that of the 2023 soil as indicated by the baseline environmental analyses (Table 4).

Figure 4 and Figure S1 show more clearly the effects of PC blend dose, W/R values, PC type, and incorporation of SC and OC on the leaching of naphthalene, a key compound in the creosote. Figure S1 shows that up to a 5 wt% difference in PC blend dose does not result in a meaningful difference in the CMR curves for a range of SVOCs including naphthalene, acenaphthene, phenanthrene and benzo(a)anthracene. This is consistent with Khuri et al. [17], who observed the same for chlorobenzene NAPL-impacted soils treated by 5 to 10 wt% of a PC blend. Figure S1 also shows that the inclusion of 1 wt% OC makes no difference for these four SVOCs common to MGP- and creosote-NAPL-impacted sites.

Figure 4A suggests that the W/R ratio had effectively no difference in the CMR plots given that CMR curves tend to be ordered (stacked) from top to bottom by total content [17], and the two CMR curves differ by a factor of 2, effectively matching their difference in total content concentrations shown in Table 4. Figure 4B shows that the cement type essentially has no influence on the naphthalene mass transfer rates, whereas Figure 4C shows that incorporating SC lowers the CMR values by approximately 50%. Figure 4D shows that the 5 and 7.5 wt% PC blend pairs with and without OC effectively overplot each other, showing little to no effect of OC in terms of leaching reductions for the 2023 sample.

Like Figure 4D, Figure 5 shows the EPA Method 1315M CMR curves for 1-methylnaphtalene, acenaphthene, phenanthrene and benzo(a)anthracene, which are indictive a wide range of hydrophobicity based on their respective octanol–water partitioning coefficient (K_ow) values. These show effectively no difference between mixes with and without OC when comparing mixes with the same PC dose. These leaching results are consistent for other VOCs/SVOCs that are prevalent at MGP- and creosote-NAPL-impacted sites [11,12,13] and for numerous pesticides treated by GAC, as summarized by Grubb and Briggs [14].

Conversely, Figure 6A shows that the use of 1 wt% OC (solid symbols) results in approximately an order-of-magnitude (10×) reduction in the EPA Method 1315M CMR curves for PCP. Figure 6B–D show that the ability of OC to reduce the leaching of chlorinated phenols is impacted by degree of chlorination, with efficacy declining rapidly to a factor of approximately two for (2,4,5-) trichlorophenol to virtually no effect for phenol (Figure 6D), just like many other VOCs/SVOCs (e.g., Figure 5 and Figure S1). Figure S2 shows a similar trend for the 2020 mixes.

Figure 1 shows the ISS footprint and distribution of PCP at the site, which consists primarily of two separate plumes. The larger of the two plumes is proximate to the creek bordering the site to the north. The 2015 study recommended the use of OC for the entire footprint at an estimated reagent unit cost of $13.37 million, but Figure 3B, Figure 5, Figure 6D, Figures S1 and S2D all clearly show that OC (the organic sorbent) has essentially no effect on the SVOCs common to creosote NAPL above conventional cement treatments, consistent with prior EPA Method 1315M studies using OC (or GAC) [11,12,13,14]. Further, no real benefit of OC for treatment of the southern PCP hotspot was anticipated to exist once the site was fully treated by ISS (and maintains K ≤ 1 × 10⁻⁶ cm/s) for the simple reason that it would take hundreds of years for PCP from this hotspot to diffuse to the creek along the northern edge of the property. The inclusion of OC was perceived to be most impactful along the northeastern perimeter of the ISS treatment zone where it would capture PCP in groundwater before it leaves the site. For this reason, the remedial design included a 50 ft (15 m) ISS treatment buffer (15,000 CY; ~11,450 m³ total) where the mix designs included 1 wt% OC, as shown in Figure 7. This targeted use of OC was deemed critical and still saved EPA approximately $12.8 million over the avoided cost of using 1 wt% OC across the entire ISS footprint.

8. Concluding Remarks

This treatability study focused on exploring if organoclay (OC) was needed as a specialty regent for the stabilization/solidification (S/S) of creosote-NAPL-impacted soils. Fourteen mix designs were prepared and tested over two events (2020, 2023) using a combination of Type I/II Portland cement (PC) or Type IL Portland–limestone cement (PLC) with slag cement (SC), bentonite (B) powder, and with and without OC. The water-to-reagent (W/R) ratio was varied between the two studies to address site mixing challenges which required a higher W/R (2.8) for bucket mixing during field piloting. Key observations are listed below:

• All mix designs easily met the site criterion of a K ≤ 1 × 10⁻⁶ cm/s.

The single most important factor impacting strength was the W/R ratio. The W/R of 2.8 used in the field pilot passed the design criteria after 28 days of curing; however, the W/R of 1.0 had a much greater strength overall, exceeding the design criteria by a factor of 5. W/R had a mixed but not a significant impact on leaching performance.

The second most important factor was the inclusion of SC to meet the unconfined compressive strength (UCS) design criteria of 50 lb/in² (345 kPa), but its inclusion did not significantly impact leaching performance.

Type IL PC performed comparable to Type I/II PC during the testing as there were no significant differences found in strength or leaching.

OC had negligible effects on UCS but offered up to a 10× reduction in PCP leaching. The impact on phenolic leaching decreased with decreasing degree of chlorination down to phenol, for which it had no discernable effect. OC had no significant impact on the leaching of other creosote constituents, consistent with the prior study and other EPA Method 1315M studies that incorporated OC and/or GAC reagents [11,12,13,14]. This is the first study that the authors are aware of that shows that OC provides significant EPA 1315M leaching reductions in hydrocarbons from monolithic, NAPL-impacted cement-treated samples.

Targeted use of OC was designated only to treat a PCP hotspot near the downgradient perimeter of ISS treatment area where it was most needed to protect a proximate stream. This focused use versus treating the entire ISS footprint with OC saved EPA approximately $12.8 million in reagent costs.