Research on the Combined Treatment of Composite Organic-Contaminated Soil Using Diversion-Type Ultra-High-Temperature Pyrolysis and Chemical Oxidation

Abstract

Remediating complex-contaminated soils demands the synergistic optimization of efficiency, cost-effectiveness, and carbon emission reduction. Currently, ultra-high-temperature thermal desorption technology is mature in terms of principle and laboratory-scale performance; however, ongoing efforts are focusing on achieving stable, efficient, controllable, and cost-optimized operation in large-scale engineering applications. To address this gap, this study aimed to (1) verify the energy efficiency and economic benefits of removing over 98% of target pollutants at a 7.5 × 10⁴ m³ contaminated site and (2) elucidate the mechanisms underlying parallel scale–technology dual-factor cost reduction and energy–carbon–cost optimization, thereby accumulating case experience and data support for large-scale engineering deployment. To achieve these objectives, a “thermal stability–chemical oxidizability” classification criterion was developed to guide a parallel remediation strategy, integrating ex situ ultra-high-temperature thermal desorption (1000 °C) with persulfate-based chemical oxidation. This strategy was implemented at a 7.5 × 10⁴ m³ large-scale site, delivering robust performance: the total petroleum hydrocarbon (TPH) and pentachlorophenol (PCP) removal efficiencies exceeded 99%, with a median removal rate of 98% for polycyclic aromatic hydrocarbons (PAHs). It also provided a critical operational example of a large-scale engineering application, demonstrating a daily treatment capacity of 987 m³, a unit remediation cost of 800 CNY·m⁻³, and energy consumption of 820 kWh·m⁻³, outperforming established benchmarks reported in the literature. A net reduction of 2.9 kilotonnes of CO₂ equivalent (kt CO₂e) in greenhouse gas emissions was achieved, which could be further enhanced with an additional 8.8 kt CO₂e by integrating a hybrid renewable energy system (70% photovoltaic–molten salt thermal storage + 30% green power). In summary, this study establishes a “high-temperature–parallel oxidation–low-carbon energy” framework for the rapid remediation of large-scale multi-contaminant sites, proposes a feasible pathway toward developing a soil carbon credit mechanism, and fills a critical gap between laboratory-scale success and large-scale engineering applications of ultra-high-temperature remediation technologies.

1. Introduction

With the deepening of industrial restructuring and the implementation of environmental protection policies, numerous chemical enterprises have been closed or relocated, leaving large-scale contaminated sites that pose a global challenge. In China, the soil pollution exceedance rate at industrial wastelands reaches 34.9% [1]. TPHs, PAHs, and PCP often co-occur as complex mixtures. The detection frequency of such mixtures in the Yangtze River Delta and Pearl River Delta industrial clusters exceeds 30% [2]. These pollutants threaten ecosystems and human health because of their carcinogenic, teratogenic, and mutagenic effects and high bioaccumulation potential [3,4,5,6], making efficient remediation imperative.

Thermal desorption (TD) has emerged as a predominant ex situ technology for the remediation of organically contaminated soil, particularly valued for its high treatment efficiency, broad applicability, and short remediation cycle, which facilitates rapid land redevelopment [7,8]. The principle of TD involves the thermal decomposition and volatilization of organic contaminants from the soil matrix. Its efficiency is predominantly governed by temperature and residence time, with higher temperatures significantly enhancing the removal of recalcitrant compounds [9,10,11]. Studies have consistently shown that increasing temperature significantly enhances the removal efficiency of recalcitrant organic compounds, such as PAHs, at ultra-high temperatures (>950 °C). Even highly stable pollutants such as PAHs can be completely mineralized via graphitization, rather than merely being transferred, thereby minimizing the risk of secondary pollution [12,13]. However, the primary drawbacks of TD, especially at ultra-high temperatures, are its substantial energy consumption, high operational costs (often exceeding 220 USD t⁻¹), and significant carbon footprint, which limit its economic and environmental sustainability for treating large soil volumes [14,15].

As an alternative or complementary approach, in situ chemical oxidation (ISCO), particularly persulfate-based advanced oxidation processes (PS-AOPs), has gained widespread attention. Persulfate activation generates powerful sulfate (SO₄•⁻) and hydroxyl (•OH) radicals that can degrade a wide range of recalcitrant organic pollutants [16,17]. PS-AOPs offer advantages such as relatively low cost, minimal energy input, and less soil disturbance [18,19]. Activation methods, including thermal, alkaline, and transition metal activation, can be tailored to specific site conditions [20,21]. Alkali activation with calcium oxide (CaO) has been shown to effectively activate persulfate while simultaneously neutralizing acidic by-products and stabilizing soil pH, which is crucial for long-term soil health [14,22,23]. Nevertheless, the effectiveness of PS-AOPs can be limited for pollutants with a high bond dissociation energy or those residing in low-permeability zones, and they may not achieve complete mineralization for all components in complex mixtures [24,25].

Given the complementary strengths and limitations of TD and PS-AOPs, their combination presents a promising strategy for managing complex organic contamination. “Segmented” or “diverted” remediation, where contaminants are directed to the most suitable treatment process based on their intrinsic properties (e.g., thermal stability and oxidizability), can potentially optimize energy–carbon–cost efficiency by avoiding over-treatment and minimizing resource waste [13,26]. This study focused on a large-scale (75,000 m³) contaminated site characterized by severe co-contamination of petroleum hydrocarbons (C10–C40 (PHCs)) (up to 7.01 × 10³ mg kg⁻¹) and PAHs such as benzo[a]pyrene (B[a]P). The site represents a typical yet challenging case, where single-technology approaches are either economically prohibitive (if using TD for the entire volume) or technically inadequate (if relying solely on ISCO for high-concentration, recalcitrant pollutants). Therefore, this site serves as an ideal platform to validate a novel diverted treatment train: directing high-thermal-stability pollutants to ultra-high-temperature (1000 °C) pyrolysis, while simultaneously treating easily oxidizable components with a CaO-activated persulfate system.

The primary objectives of this engineering application research are as follows: (1) to verify the field-scale efficacy (>98% removal) and operational stability of this combined technology for a 75,000 m³ project; (2) to quantitatively elucidate the energy–carbon–cost optimization mechanism achieved with this parallel, technology-based diversion strategy; and (3) to accumulate comprehensive case-specific data and experience to support the standardized, large-scale application of such integrated remediation solutions.

2. Materials and Methods

2.1. Site Overview

The study site was located on the Yellow River terrace in the Guanzhong Basin, characterized by 0–3.5 m of artificial fill, 3.5–9.0 m of silty clay, and 9.0–18.0 m of fine sand. The permeability coefficient of the fine sand layer was 1.2 ± 0.3 m d⁻¹ (pumping test; n = 16), and the hydraulic gradient was approximately 2.3‰. The groundwater table was buried at a depth of 12–17.7 m (measured in spring 2024, with an annual variation of approximately 1 m), which was deeper than the maximum excavation face (7 m). Thus, the remediation operation was not affected by groundwater buoyancy or seepage scouring. From 2010 to 2020, the site operated a 600,000-ton-per-year coal-to-methanol unit, which produced by-products such as sulfur and argon, along with coal tar and methanol oil fractions. There was a risk of polycyclic aromatic hydrocarbon (PAH) and petroleum hydrocarbon leakage during storage, transportation, and the “three wastes” (waste gas, wastewater, and solid waste) treatment processes.

2.1.1. Pollution Identification

Five pollutants exceeding the standards were detected in the 0–7 m soil layer (

Table 1. Analysis and statistics of soil pollution detection results of the site (unit: mg kg⁻¹).

Test Items	PHC	B[a]P	D[ah]A	B[a]A	PCP
Screening values *	4500	1.5	1.5	15	2.7
Control values *	9000	5.5	5.5	55	/
Maximum	7010	3.91	2.92	22.4	4.45
Detection rate (%)	96.8	6.5	1.1	7.7	37.1
Maximum content (point-depth)	S14-0.5	XS09-0.5 m	XS09-0.5 m	XS09-0.5 m	XS26-2.0
Maximum over-limit multiple	0.56	1.61	0.95	0.49	0.65

Screening values *; control values *. PHC (C₁₀–C₄₀) data source: Supplementary Indicators for Screening values of Soil pollution Risk Control for Construction Land in Shanghai (Trial, 2021)–Class I Land Use Standards; B[a]P, D[ah]A, B[a]A, and PCP data all refer to GB 36600-2018 [27] (Class I land).

). Petroleum hydrocarbons (C₁₀–C₄₀ (PHCs)) had a maximum concentration of 7.01 × 10³ mg kg⁻¹, corresponding to an exceedance factor of 0.56 (relative to the screening value). The maximum concentrations of B[a]P, dibenzo[ah]anthracene (D[ah]A), benzo[a]anthracene (B[a]A), and PCP were 3.91, 2.92, 22.4, and 4.45 mg kg⁻¹, respectively, with corresponding exceedance factors of 1.61, 0.95, 0.49, and 0.65 (the screening values are presented in Table 1). Concentrations exceeding the standards fell between screening and control levels. High-value areas were concentrated in the slurry preparation area, diesel tank gasification area, coal storage silos, and wastewater unit, showing spatial coupling with the production units. Based on the USEPA RBCA model, the combined carcinogenic risk (incremental lifetime cancer risk (ILCR)) for on-site workers was 1.1 × 10⁻³, and the ILCR for future residential scenarios was 1.2 × 10⁻⁴, both exceeding the acceptable level of 1 × 10⁻⁵. Therefore, active remediation was selected to reduce the risk.

2.1.2. Demarcation of Pollution Remediation Areas

The pollution characteristics of different areas were identified via high-density environmental investigation and risk assessment. Subsequently, different remediation units were delineated, and differentiated remediation technologies and targets were established. The remediation zones were divided into 11 units (zones A–L) based on the planar distribution of pollutant exceedances. Vertically, the area was divided into four layers—0–1.5 m, 1.5–3.5 m, 3.5–5.0 m, and 5.0–7.0 m—according to concentration attenuation characteristics (Figure 1), with a total earthwork volume of approximately 7.5 × 10⁴ m³. The excavation sequence followed the principle of “from far to near”. After the acceptance of each section, clean soil was backfilled, and exposed surfaces were immediately sealed to reduce the risk of secondary diffusion.

2.2. Determination of Process Parameters and Equipment

2.2.1. Split-Flow Process

Based on the pollution spectrum (high-ring PAHs, PHCs, and PCP) and site constraints (e.g., excavation depth ≤ 7 m; groundwater level > 12 m), a multi-criteria screening approach was employed to develop an optimized coupled remediation system for segmented ex situ thermal desorption (ESTD) and in situ chemical oxidation (ISCO). This system was designed based on the synergistic evaluation of thermal stability and chemical oxidizability [28,29]. The segmentation decision was based on two core criteria: the T_90% thermal stability index of pollutants and the kinetic parameters of radical reactions. Using thermogravimetric analysis (TGA) and reaction kinetics assessment, pollutants satisfying T_90% > 550 °C or k_SO4•⁻ < 5 × 10⁷ M⁻¹ s⁻¹ were classified as the “high-thermal-stability type”, mainly comprising long-chain PHCs (C₁₅–C₄₀) and chlorinated organic compounds [30]. The remaining pollutants were categorized as “oxidizable” [31], typified by PAHs.

The sequence of soil excavation and the areas and volumes for the layered remediation of various soil pollutants were designed based on the preliminary and detailed investigation results of 1636 sets of data from contaminated sites, as well as the soil’s spatial location information. Take area A in Figure 1 as an example (Figure 2;

Table 2. Stratified soil excavation statistics (taking area A as an example) (unit: m³).

Category	Environmental Assessment No.	Excavation No.	Pollution Type	Theoretical Volume	Actual Volume	Transport Route	Treatment Measure
Polluted soil excavation	A1	A1-1	PAHs	1211.23	1209.84	A foundation pit–pretreatment workshop	Chemical oxidation
	A2	A1-2	PCP	1255.81	2902.97	A foundation pit–pretreatment workshop	Thermal desorption
	A3	A1-3	Organic complex pollution	1578.66		A foundation pit–pretreatment workshop
Staggered soil	QA1-1	—	—	553.97	552.25	Transported to clean soil temporary storage area	To be tested
Total				4599.67	4665.06

).

Contaminated soil was excavated using a stratified and zoned refined operation strategy to initially separate and transport the two contaminated soil types to the pretreatment workshop. High-concentration, high-thermal-stability contaminated soil (approximately 6.4 × 10⁴ m³) was immediately transported to the ESTD pretreatment workshop, while medium- and low-concentration, easily oxidizable contaminated soil (approximately 1.1 × 10⁴ m³) was conveyed to the alkali-activated persulfate ISCO treatment area (Figure 3). The total volume ratio of zoned remediation was approximately 85:15. Subsequent synergistic assessments of energy consumption, carbon emissions, and cost were conducted based on this demarcation.

2.2.2. ESTD Process Parameters and Equipment

A 3.6 × 10³ m² (20 cm thick) C30 concrete-hardened ESTD operation area was constructed within the site, equipped with rotary kilns. Kilns nos. 1 and 2 had rated feed rates of 35–45 t·h⁻¹ and 10–20 t·h⁻¹, respectively. The two kilns could be started and stopped independently. Contaminated soil was screened and crushed to ≤50 mm and adjusted for moisture content. When the moisture content exceeded 20 wt%, quicklime was added to adjust it to approximately 20 wt% (in compliance with domestic and international ESTD engineering practices and the technical specifications of HJ 25.5-2018 [32]) for pretreatment [33]. Natural gas was injected via burners, with 2000 Nm³·h⁻¹ for burner no. 1 and ≥800 Nm³·h⁻¹ for burner no. 2 (this study relied on the numerical data from the summary report of the construction source).

The thermal desorption remediation experiment for B[a]P that contaminated the soil showed that at a temperature of 900 °C, the removal rate of B[a]P was approximately 90% to 92%, and the residual concentration still reached 0.3 to 0.5 mg/kg. When the temperature rose to 1000 °C, the removal rate of B[a]P increased to 95% to 97%, and the residual concentration dropped to 0.1 to 0.2 mg/kg (compliant with the standard). The removal rate only increased by 0.8% when further increasing the temperature to 1050 °C [34,35]. High-temperature flue gas came into direct contact with the material, providing energy for the thermal decomposition and mineralization of organic matter (typified by PHCs C₁₅–C₄₀ and PCP). For the heating process, temperatures were controlled within the range from 100 °C in the inlet drying section to 1000 °C in the high-temperature cracking/mineralization section, then to 900 °C in the medium-temperature transition section, and, finally, to 500 °C in the outlet cooling section. The total residence time of the soil for pyrolysis and mineralization was 30 min [28,34,35,36].

The specific process flow was as follows: contaminated soil → feeding equipment → rotary kiln thermal desorption equipment → discharge equipment → treated soil → cyclone separator → secondary combustion chamber → quench tower → bag filter → induced draft fan → acid removal device → chimney (the process flow is illustrated in Figure 4).

2.2.3. ISCO Process Parameters and Reagents

The oxidation workshop consisted of a 2400 m² color-steel shed and a 20 cm thick C30 concrete base plate, with rainproof and load-bearing functions. Medium- and low-concentration, easily oxidizable contaminated soil was transported to the ISCO treatment area for alkali-activated persulfate processing. ISCO employs the CaO–Na₂S₂O₈ system as follows: 2–4 wt% CaO is used to provide a pH 10–12 alkaline environment for persulfate activation, offering advantages including pH regulation, thermal activation, and cost control; Na₂S₂O₈ (4–8 wt%) acts as the oxidant [37,38], with its theoretical oxidation equivalent determined based on the total oxygen demand (TOD) of the target pollutant. A 1.5-fold safety factor was incorporated to ensure complete reaction [22]. In this study, the dosage of the alkaline activator (calcium oxide (CaO)) was 3.66 wt%, corresponding to a total of 364 t; the dosage of the oxidant (sodium persulfate (Na₂S₂O₈)) was 6.34 wt%, totaling 701 t.

The contaminated soil was pretreated using a screening and crushing hopper (ALLU SMH 3-23 (ALLU, Lahti, Finland); sieve aperture ≤ 50 mm; operated with a CAT320 excavator (Caterpillar, Peoria, IL, USA). Its moisture content was adjusted to 30–40% via a spiral spray system for quantitative water replenishment, employing a closed-loop system of “variable-frequency spiral + fan-shaped nozzle + load cell”. The reagent was effectively diffused in the soil pore water. A rotary tiller-turning machine was used to apply the reagent evenly. After mixing, the pile was covered with a film and cured at 25 °C, with a typical curing period of 3 days and a dynamic adjustment range of 2–5 days [39]. For endpoint determination, ORP ≥ +450 mV (Pt electrode; 25 °C; SL 1:2.5) and residual S₂O₈²⁻ ≤ 5 g kg⁻¹ [40]. The daily processing capacity of a single equipment set was 350 m³. The process parameters and endpoint determination collectively ensured that the pollutant removal rate exceeded 98% and secondary risks were controllable. The process flow is illustrated in Figure 5.

2.2.4. Remediation Effect Evaluation Parameters

(1)

Removal rate

The removal efficiency (RE) was calculated based on the 95% upper confidence limit (UCL) of the pre- and post-treatment concentrations to quantify the overall reduction in contaminants achieved by the combined ultra-high-temperature pyrolysis and chemical oxidation process. Using the 95% UCL instead of the arithmetic mean accounted for the spatial heterogeneity of the pollutant distribution and guaranteed that the calculated reduction protected the worst-case hotspots within each assessment unit [41,42]. The RE was then expressed as the percentage decrease in this conservative concentration estimate, as given using Equation (1):

RE (%) = [(C_{0, 95%UCL} − C_{after, 95%UCL})/C_{0, 95%UCL}] × 100

(1)

where RE represents the removal rate; C_{0, 95%UCL} represents the 95% upper confidence limit of the contaminant concentration before remediation; and C_{after, 95%UCL} represents the 95% upper confidence limit of the pollutant concentration after remediation.

(2)

Compliance rate

While the removal rate documents the magnitude of concentration decline, the compliance rate evaluates whether the remediation objective has been met in every spatial sub-domain. An assessment unit is, therefore, declared “compliant” only if the concentration at all sampling points inside it is ≤ the remediation target value. This point-by-point criterion prevents the false-positive conclusion of “average compliance” that can arise when a mean or UCL value masks isolated but potentially risky exceedances [41,43]. The site-wide compliance rate is subsequently computed with Equation (2):

Compliance rate (%) = (number of compliance assessment units/total number of assessment units) × 100%

(2)

Note: An assessment unit is considered “compliant” if the concentration at all sampling points within it is ≤ the remediation target value. This avoids the risk of partial non-compliance being masked by “average compliance”.

(3)

Carbon emission accounting method

Carbon accounting was conducted using the emission factor method to quantitatively evaluate the carbon emission reduction efficiency of the split-type ultra-high-temperature pyrolysis system. The system boundary was defined as the direct and indirect carbon emissions generated from energy consumption during the pyrolysis process, the primary source of greenhouse gas emissions throughout the remediation period. A reference scenario was established to represent the conventional energy structure for thermal desorption remediation in China, which consists of grid electricity supplemented with coal-fired boiler heating. Carbon emissions from this reference scenario were compared with the actual carbon emissions from on-site natural gas-based heating, where emission factors were sourced from the latest Chinese Life Cycle Database and IPCC guidelines. Additionally, based on a forward-looking low-carbon energy scenario, a zero-carbon heat source configuration (70% solar-molten salt energy storage + 30% green electricity procurement) was assumed to reduce the carbon factor, and the corresponding carbon dioxide equivalent (CO₂e) was calculated.

2.3. Sampling and Testing Methods

2.3.1. Sampling Protocol

The sampling density and workload at each stage were determined based on HJ25.5-2018, employing a 40 m × 40 m systematic grid or densification for units of ≤500 m³ (

Table 3. Summary of sampling points for restoration effect evaluation.

Stages	Region	Unit Volume/Area	Main Sample Number	Parallel Sample	Little Calculator
Clearing effect	Bottom of the pit	≤40 m × 40 m grid	61	10	71
	Side wall a	Layering	200	28	228
After repair	Chemical oxidation reactor	≤500 m3	25	5	30
	Thermal desorption reactor	≤500 m3	152	20	172
	Secondary pollution zone	Grid ≤ 40 m × 40 m	84	12	96
Combined			548	79	627

^a: vertical stratified sampling when the lateral wall depth is greater than 1 m. The first layer is topsoil (0–0.2 m depth). Below a 0.2 m depth, each 1–3 m layer is separated, and less than 1 m is combined with the previous layer.

).

In terms of quality control, 79 groups of on-site parallel samples accounted for 14.4% (>10%). For every 20 samples, 1 CRM (certified reference material) was inserted, with a relative error of ≤±10%, except for dioxin at ±4%. Seven sets of VOCs (volatile organic compounds) were used in the full procedure, alongside seven sets of transport blanks. The spiked recovery rates were 70–130%, all meeting the general QC (quality control) requirements of HJ 168-2020 [44].

2.3.2. Sample Preservation

During sample preservation and transportation, all samples were refrigerated at 4 °C, stored and transported under light-protected conditions, and promptly delivered to the laboratory for analysis (

Table 4. List of soil sample storage specifications.

Target Substances	Container	Filling Method	Storage Temperature	Shelf Life	Notes
Petroleum hydrocarbons	250 mL glass	Headspace minimum	4 °C	14 d	—
PAHs, pentachlorophenol	250 mL glass	Headspace minimum	4 °C	10 d	—
Benzene series (low concentrations)	40 mL brown	2 × 40 mL + dry weight	4 °C	7 d	Add a stirring bar
Benzene series (high concentration)	40 mL brown	Methanol protection	4 °C	14 d	Extract within 7 days

).

2.3.3. Detection Methods

The target pollutants —PCP, 16 PAHs, and PHCs (C₁₀–C₄₀)—in the soil were detected per the following national standards: for PCP, the Determination of Phenolic Compounds in Soil and Sediments–Gas Chromatography Method (HJ 703-2014) [45]; for the 16 PAHs, the Soil and Sediment–Determination of Semi-Volatile Organic Compounds–Gas Chromatography–Mass Spectrometry Method (HJ 834-2017) [46]; and for PHCs (C₁₀–C₄₀), the Soil and Sediment–Determination of Petroleum Hydrocarbons (C₁₀–C₄₀)–Gas Chromatography Method (HJ 1021-2019) [47].

Extraction, purification, and instrumental conditions were performed following the provisions of these standards. The details are briefly described below.

PCP was analyzed via gas chromatography with a micro-electron capture detector (GC-μECD) following acetylated derivatization. The method detection limit (MDL) was 0.03 mg kg⁻¹, and the recovery rates ranged from 85 to 128% (n = 79; median = 96%). One low-concentration sample exhibited a recovery rate of 42.6% (below 70%), which was confirmed via re-derivatization/extraction and replaced with remeasured values. The relative standard deviation (RSD) was ≤6%.

The 16 PAHs were analyzed via gas chromatography–mass spectrometry in the selected ion monitoring mode (GC-MS-SIM) using isotope internal standards. The MDL ranged from 0.02 to 0.09 mg kg⁻¹, with recovery rates from 87 to 121%, and an RSD of ≤8%.

The PHCs (C10–C40) were analyzed via gas chromatography with a flame ionization detector (GC-FID) following silica gel purification. Piecewise integration was performed for the C₁₀–C₂₅ and C₂₆–C40 fractions. The MDL ranged from 2 to 3 mg kg⁻¹, with recovery rates ranging from 85 to 109% (n= 79; median = 98%). One low-concentration sample exhibited a recovery rate of 61.2% (below 70%), which was confirmed via re-derivatization/extraction and replaced with remeasured values. The RSD was ≤5%.

3. Results

3.1. Results of Monitoring After Excavation Pit Completion

After the excavation pit was completed, a total of 261 samples (including those from the pit bottom and side walls) were collected, and all monitoring points met the remediation target requirements. PCP was not detected in any sample (detection limit: <0.07 mg kg⁻¹), resulting in a 100% removal rate. The PHCs (C₁₀–C₄₀) had a residual concentration of 36.8 ± 19.5 mg kg⁻¹, equivalent to 7% of the USEPA regional screening level (RSL), with a removal rate (RE) of >99%. The removal rate of B[a]A reached 99.3%. For B[a]P, the removal rates were 91.3% and 92.2% in samples from the side wall and pit bottom, with average residual concentrations of 0.19 ± 0.08 mg kg⁻¹ and 0.17 ± 0.07 mg kg⁻¹, respectively. Both residual concentrations were significantly lower than the screening value of 1.5 mg kg⁻¹, with a 100% compliance rate (

Table 5. Statistics of residual pollutants after foundation pit excavation.

Indicators	PHC		PCP		B[a]A		B[a]P		D[ah]A
	Side Wall	Pit Bottom	Side Wall	Pit Bottom	Side Wall	Pit Bottom	Side Wall	Pit Bottom	Side Wall	Pit Bottom
Average (mg kg−1)	42.8	30.7	-	-	0.15	0.15	0.19	0.17	-	0.1
Standard deviation	22.0	16.8	-	-	0.05	0.05	0.08	0.07	-	0.0
95% UCL	48.5	34.9	-	-	0.17	0.17	0.22	0.19	-	0.10
Removal rate (%)	99.2	99.4	100	100	99.3	99.3	91.3	92.2	100	96.6
Compliance rate (%)	100	100	100	100	100	100	100	100	100	100

Table notes: Side wall and pit bottom refer to the side wall and the bottom of the excavation pit for contaminated soil removal, respectively.

).

3.2. Remediation Results of Contaminated Soil

3.2.1. Remediation Effect of Enhanced Thermal Desorption (ESTD)

The overall compliance rate was 100% for the remediation effect of thermal desorption (n = 152) (

Table 6. Statistics of residual pollutants in thermal desorption piles and secondary pollution areas.

Indicators	PHC		PCP		B[a]A		B[a]P		D[ah]A
	ESTD	SPA	ESTD	SPA	ESTD	SPA	ESTD	SPA	ESTD	SPA
Average (mg kg−1)	38	36.8	—	0.01	0.14	0.02	0.14	0.03	0.01	<0.01
Standard deviation	19.1	18.7	—	0.063	0.05	0.103	0.07	0.119	0.03	0.05
95% UCL	42.82	44.87	—	0.05	0.06	0.08	0.1	0.09	0.00	0.02
Removal rate (%)	99.3	99.3	—	99.7	99.4	99.9	93.5	98.6	99.7	99.7
Compliance rate	100		100			100		100

Table notes: ESTD: detection results of residual pollutants after ex situ thermal desorption (ESTD) remediation; SPA: detection results of pollutants in the secondary pollution area (SPA).

). The average residual concentrations were 38.0 ± 19.1 mg·kg⁻¹ for petroleum hydrocarbons (PHCs), 0.14 ± 0.05 mg·kg⁻¹ for benzo[a]anthracene (B[a]A), and 0.01 ± 0.03 mg·kg⁻¹ for dibenzo[a,h]anthracene (D[ah]A), with removal rates exceeding 99% for all three contaminants. Pentachlorophenol (PCP) was not detected (detection limit: <0.07 mg·kg⁻¹), indicating that chlorinated organic compounds were completely mineralized at temperatures of 900–1000 °C. The average residual concentration of benzo[a]pyrene (B[a]P) was 0.14 ± 0.07 mg·kg⁻¹, corresponding to a removal rate of 93.5%. This residual level was close to the upper limit of the technology’s high-temperature performance [48]. A risk assessment was conducted, which showed that the 95% upper confidence limit (UCL) of the B[a]P residual concentration was 0.21 mg·kg⁻¹. The incremental carcinogenic risk calculated using the regional screening level (RSL) model was <1 × 10⁻⁶, which fell within the acceptable risk level specified by the United States Environmental Protection Agency (USEPA). In conclusion, the residual risk of soil after thermal desorption remediation was acceptable, and the treated soil could be safely backfilled for subsequent reuse.

3.2.2. Results of Chemical Oxidation Stage

The initial concentrations of B[a]A and B[a]P in the soil were 22.4 mg·kg⁻¹ and 3.91 mg·kg⁻¹, respectively. After treatment, the mean residual concentrations decreased to 0.08 ± 0.084 mg·kg⁻¹ (99.6% removal rate) for B[a]A and 0.12 ± 0.103 mg·kg⁻¹ (94.5% removal rate) for B[a]P. The 95% upper confidence limits (UCLs) of the endpoint concentrations were 0.21 mg·kg⁻¹ (B[a]A) and 0.29 mg·kg⁻¹ (B[a]P), respectively, equivalent to only 1.4% and 19.3% of the screening values(

Table 7. Statistics of sampling data for heterotopic chemical oxidation restoration reactor.

Indicators	Average (mg kg−1)	Standard Deviation	95% UCL	Removal Rate (%)	Compliance Rate (%)
B[a]A	0.08	0.084	0.21	99.6	100
B[a]P	0.12	0.103	0.29	94.5	100
D[a,h]A	—	—	—	100	100
Benzene and benzene compounds	—	—	—	100	100

). Both residual concentrations were below the limits specified in the Chinese National Standard (GB) 36600-2018 [27]. High-ring PAHs were not detected (method detection limit (MDL) = 0.1 mg·kg⁻¹), indicating that free radicals exhibit a highly efficient degradation capacity for PAHs with ≥4 rings. In this study, alkali-activated persulfate treatment for 3 days achieved a 94.5% removal rate of B[a]P. This rate was higher than the 85.41% removal rate reported in the literature for the used a mixed oxidant (persulfate + calcium peroxide) [49,50]. This finding provides additional evidence that “alkaline activation conditions enhance the degradation capacity of high-ring PAHs” [51,52]. As shown in Table 6, the remediation results demonstrate that no additional engineering measures are required under the current remediation targets and sampling density.

Soil samples were collected from four locations within the excavated soil pile of pit E (where the concentrations of PAHs and petroleum hydrocarbons (C₁₀–C₄₀) were the highest and the pollution scope was the most extensive) to verify whether new secondary compounds were generated in the soil during the chemical oxidation process. Comprehensive scanning of volatile organic compounds (VOCs) and semi-volatile organic compounds (SVOCs) was conducted. Based on the comprehensive scanning results, no additional substances were detected. Based on the comprehensive scanning results, no additional secondary compounds were detected.

Inferential statistics were added to rigorously test the significance of inter-process differences. An independent-samples t-test (ESTD vs. ISCO; n₁ = n₂ = 7) revealed no statistically significant difference in the benzo[a]pyrene removal efficiency (t(12) = −0.46; two-tailed p = 0.65; and Cohen’s d = 0.25). Both stages exceeded 92% removal, verifying process reliability.

3.3. Test Results for Secondary Contaminated Areas

Ninety soil samples were collected from potential secondary contaminated areas for analysis to assess the remediation effect (Table 6). These areas included the remediation greenhouse, the construction waste washing area, the sewage treatment area, on-site roads, the area awaiting inspection, and the temporary storage area for non-contaminated soil. The results showed the following: For PHCs, the average residual concentration was 36.8 ± 18.7 mg·kg⁻¹, the 95% upper confidence limit (UCL) was 44.87 mg·kg⁻¹ (equivalent to 0.85% of the target value), and the removal rate reached 99.3%. B[a]A had a detection rate of only 4.4%, with an average residual concentration of 0.020 ± 0.103 mg·kg⁻¹. B[a]P had a detection rate of 7.8%, with an average residual concentration of 0.030 ± 0.119 mg·kg⁻¹. D[ah]A had a detection rate of 3.3%, with all detected concentrations < 0.01 mg·kg⁻¹. PCP reached 99.7%. This variation in detection rates and residual concentrations reflects the spatial heterogeneity of pollutant distribution, which is consistent with the typical characteristics of remediated contaminated sites. Over 92% of the samples had undetectable levels of target pollutants; only a small number of sites exhibited trace residues, although this resulted in a higher standard deviation (coefficient of variation (CV) > 1.0), but the detected residue concentrations were far below the screening values, thus not affecting the overall effectiveness of the remediation [53,54].

No significant signs of pollutant rebound or cross-contamination were observed during the monitoring period. No additional engineering measures are required under the current remediation targets. A long-term monitoring plan should be established to regularly evaluate the sustainability of the remediation effect.

4. Discussion

4.1. Removal Efficiency and Contaminant–Technology Matching Patterns

The ex situ thermal desorption–chemical oxidation combined technology exhibited excellent removal performance for the five target pollutants (Figure 6a–e).

As shown in Figure 6a–e, the removal rates of the five target pollutants ranged from a minimum of 91.3% (for B[a]P, due to its more stable molecular structure) to over 98% for the other four pollutants. These rates meet the “excellent” standard (removal rate ≥ 90%) specified in the Catalogue of Contaminated Site Remediation Technologies. This confirms that the ex situ thermal desorption–chemical oxidation combined technology is highly efficient, stable, and targeted for remediating complex organically contaminated soil at chemical sites. The contaminant–technology coupling mechanism can be categorized into three typical groups based on the matching between the physicochemical properties of pollutants and the remediation technology. This classification system provides important theoretical guidance for the precise remediation of complex contaminated sites.

(1)

High-Thermal Sensitivity/Oxidizability Group

PHCs and PCP showed excellent thermal desorption performance at 1000 °C, with removal rates of 99.3% and >99%, respectively. Their residual concentrations accounted for only 0.9% and 0.4% of the corresponding screening values. This efficient removal was mainly attributed to the thermodynamic properties of these two pollutants. For PHC, thermal cracking occurs primarily in long-chain alkane components (C₂₀–C₄₀), with a C–C bond homolytic cleavage energy ranging from 350 to 400 kJ/mol. For PCP, the C–Cl bond energy is approximately 330 kJ/mol. As calculated using the Arrhenius equation, the cracking reaction rate constant for these pollutants could reach 10⁶–10⁷ s⁻¹ at 1000 °C, and a residence time ensured complete conversion [50]. Additionally, three factors further enhanced the removal efficiency: 1. the rapid volatilization of HCl at high temperatures reduced chlorine sources; 2. the sufficient oxygen supply promoted complete oxidation; and 3. the short residence time inhibited the formation of secondary reactions. Consequently, the dechlorination process effectively controlled the generation of toxic intermediates, such as dioxins [55].

(2)

Refractory Polycyclic Aromatics

The removal rates of three PAHs—B[a]A, D[ah]A, and B[a]P—were 99.4%, 99.7%, and 93.5%, respectively, with B[a]P showing the lowest rate. According to thermal desorption kinetic theory, the PAH removal efficiency is closely related to molecular weight, ring number, and substituent type. B[a]P failed to fully volatilize during the 900 °C residence period due to its high octanol–water partition coefficient (Kow) and aging effects [12,13], which explains its relatively low thermal desorption efficiency. When the treatment temperature was increased from 900 °C to 1000 °C, the removal rate of B[a]P entered the 93–97% plateau range reported in the literature, with a marginal improvement of <1% [15,28]. This indicates that the process reached its thermodynamic limit [33,34]. This conclusion is verified by the engineering practice results of this study.

(3)

Oxidation-Specific Group

Alkali-activated persulfate exhibited excellent selective degradation capacity for PAHs, achieving removal rates of 98.6–99.7%. PAH molecules are rich in π-electrons, making them susceptible to electrophilic attack by free radicals. Sulfate radicals (SO₄•⁻) primarily attack PAHs via an electron transfer mechanism. Hydroxyl radicals (•OH) act through addition and hydrogen abstraction mechanisms. Studies have shown that the pH value, as a non-negligible factor, can affect the generation of active free radical species. The generation of SO₄•⁻ decreases because of SO₄•⁻ scavenging under acidic conditions [56,57]. Under alkaline conditions (pH > 10), the conversion of SO₄•⁻ to •OH becomes increasingly important in the degradation process; thus, alkaline conditions should predominate over acidic conditions in a treatment system [58,59]. Additionally, SO₄•⁻ has higher selectivity than •OH, which promotes the decomposition of organic radicals and ensures efficient degradation. The alkali activation process involves a complex chain reaction, with the core step being OH⁻-catalyzed decomposition of persulfate to generate strongly oxidizing SO₄•⁻ and •OH (Equation (3)) [50]:

S₂O₈²⁻ + H₂O (catalyzed by OH⁻) → SO₄²⁻ + SO₄•⁻ + 2OH⁻
SO₄•⁻ + OH⁻ → SO₄²⁻ + •OH             

(3)

The high pH environment inhibited chloride ion oxidation and reduced the formation of chlorinated by-products [60]. The absence of secondary chlorinated products in this study confirmed this observation. However, uncertainties remain regarding the intermediate products of the oxidation process, which require further monitoring.

Ultra-high-temperature pyrolysis (900–1000 °C) inevitably converts soil into an engineered substrate rather than a biologically active medium. The soil organic matter content decreases, pH values shift to predominantly alkaline levels, the microbial biomass becomes undetectable, and the enzyme activity is nearly zero [57,58]. These characteristics render the treated material non-compliant with ecological reuse standards, but it is suitable for industrial hardened flooring where no ecological receptors are present. If green spaces or agricultural functions are planned for the site in the future, ultra-high-temperature remediation should be avoided; alternatively, post-remediation engineered soil reconstruction (including compost addition, pH adjustment, and biological proliferation) should be implemented.

4.2. Time Efficiency: Synergistic Optimization of Microscopic Kinetics and Macroscopic Scheduling

4.2.1. Microscopic Kinetics Analysis

The daily processing capacity of the parallel system reached 987 m³·d⁻¹ (95% confidence interval (CI): 945–1029 m³·d⁻¹), which far exceeded the average processing rate of 125.75 ± 96.69 m³·d⁻¹ from 18 cases in the literature (Appendix A 

Table A1. Summary Table of Remediation Cases of Contaminated Soil.

Number	Project Location	Technical Route	Scale m3	Daily Processing Volume m3 d−1	Energy Consumption kWh m−3	Unit Cost USD m−3 *
1	A pesticide factory in Ningbo	Indirect thermal desorption (gas, 950 °C)	24,808	120	763	106.31
2	Northeast Pharmaceutical Group	Direct + indirect thermal desorption (natural gas, 1000 °C)	51,078.5	240	820	139.28
3	A certain steel plant in Beijing	Reverse burning rotary kiln (1000 °C)	10,000	150	900	139.28
4	Suzhou A Chemical Plant	Resistance heating (ERH, 950 °C)	272,214	300	940	132.31
5	Suzhou B Chemical Plant	Resistance heating (ERH, 950 °C)	403,097	320	960	135.10
6	Nanjing C Chemical Plant	Heat conduction heating (TCH, electric, 900 °C)	36,767	110	980	137.88
7	A chemical site in Guangzhou	Heat conduction heating (TCH, gas, 900 °C)	24,864	130	870	128.13
8	Tianjin D Chemical Plant	Heat conduction heating (TCH, electric, 900 °C)	29,391	100	990	140.67
9	Wuhan E Chemical Factory	Heat conduction heating (TCH, electric, 950 °C)	66,000	140	970	136.49
10	Shanghai F Chemical Plant	Heat conduction heating (TCH, gas, 900 °C)	300	15	890	129.53
11	Anchorage (AK, USA)	Heat conduction heating (TCH, electric, 900 °C)	1529	22	464	167.13
12	A decommissioned solvent plant in Jiangsu	Gas thermal desorption (950 °C, 99.8% removal rate)	5000	50	1010	153.20
13	A pesticide factory in Beijing	Thermal desorption (940 °C)	3000	40	1050	160.17
14	A pesticide factory in Tianjin	Thermal desorption (900 °C, 30 min)	2000	30	1100	167.13
15	Organophosphorus pesticide factory in Sichuan	Thermal desorption (900–950 °C)	4000	45	1050	154.60
16	A coking plant in Anhui Province	Thermal desorption (1000 °C, 30 min)	56,000	200	900	111.42
17	A steel plant in Chongqing	Chemical oxidation + reactor thermal desorption	3483	80	1346	167.13
18	Pringy, France	In situ thermal desorption (electric, 80 °C)	10,000	30	4860	348.19

* The calculation was based on the average exchange rate of the RMB against the US dollar in October 2025: 1 US dollar ≈ 7.18 RMB.

). This demonstrates a significant time efficiency advantage, attributed to the synergistic optimization strategy of microscopic reaction acceleration and macroscopic process parallelization. For organic cracking reactions with an apparent activation energy (E_a) of approximately 200 kJ·mol⁻¹, the Arrhenius equation (Equation (4)) was used to quantify the effect of temperature on the reaction rate constant (k):

k = A × exp(−E_a/RT)
E_a = 200 kJ/mol, R = 8.314 J/(mol·K), T₁ = 1173 K (900 °C), and T₂ = 1273 K (1000 °C).
k₂/k₁ = exp[E_a/R × (1/T₁ − 1/T₂)] = exp[200,000/8.314 × (1/1173 − 1/1273)] ≈ 2.8

(4)

This calculation confirms that increasing the temperature from 900 °C to 1000 °C raises the reaction rate constant by 2–3-fold. This result is fully consistent with the 2–3-fold increase reported in the literature [61], verifying the scientific validity of the temperature optimization strategy. The elevated reaction rate constant reduced the soil residence time in the rotary kiln from the reference value of 45 min to 30 min, directly cutting the critical path time by 33% while still meeting the requirements for complete pollutant conversion [62]. Key chemical reactions during this process included

Thermal decomposition of n-eicosane: C₂₀H₄₂ + (61/2)O₂ → 20CO₂ + 21H₂O
Oxidative decomposition of pentachlorophenol (PCP): C₆HCl₅O + (11/2)O₂ + H₂O → 6CO₂ + 5HCl
Thermal oxidation of anthracene: C₁₄H₁₀ + (33/2)O₂ → 14CO₂ + 5H₂O

(5)

The temperature increase strategy enhanced the processing capacity of a single unit while also significantly reducing the per-unit energy consumption and carbon emission baseline. This underscores the engineering significance of residence time optimization.

4.2.2. Macro Process Scheduling Optimization

Traditional “serial” process configurations for thermal desorption and chemical oxidation units involve significant process waiting times. To address this, this study proposed a parallel diversion strategy based on real-time matching of “pollutant characteristics–treatment pathway” (Figure 3), wherein 1. contaminated soil was rapidly screened; 2. highly thermally sensitive fractions were directly fed into the 1000 °C rotary kiln; and 3. refractory PAHs were simultaneously directed to the alkali-activated persulfate oxidation stockpiles. The two units operated independently at full load, with no intermediate buffer period, thus eliminating process gaps. Using United States Environmental Protection Agency (USEPA) standard parameters (slope factor (SF) = 3.85 mg·kg⁻¹·d⁻¹; exposure frequency (EF) = 350 d·y⁻¹; exposure duration (ED) = 30 a; body weight (BW) = 70 kg; average time (AT) = 80 a; and concentration difference (ΔC) = 0.12 mg·kg⁻¹), this scheduling mode shortened the on-site exposure window by approximately 30 days. The calculated incremental lifetime cancer risk (ΔILCR ≈ 3 × 10⁻⁵) quantitatively confirms the marginal benefits of macroscopic parallel scheduling. It compresses remediation cycles and reduces population health risks.

4.3. Comprehensive Assessment of Carbon–Energy–Cost Synergies

4.3.1. Synergistic Mechanisms for Energy Efficiency Improvement

At the microscopic level, the Arrhenius effect at 1000 °C increased the cracking rate constant by 2.8-fold, shortening the soil residence time in the rotary kiln from 45 min to 30 min and directly reducing the heating load by 33%. Additionally, the high-temperature environment ensured complete oxidation in a single pass, eliminating energy consumption associated with reprocessing. At the macroscopic level, modular waste heat recovery systems cooled the flue gas from 500 °C to 200 °C. Following the principle of “high-grade heat for high-demand uses, low-grade heat for preheating,” this recovered heat was fed stepwise into the front-end drying and back-end hot water systems, enabling directed conversion of waste heat into usable energy. An intelligent temperature control algorithm dynamically matched load fluctuations and immediately redistributed excess heat, enhancing the system’s thermal efficiency. This study integrated and coupled microscopic reaction thermodynamics with macroscopic energy systems at an engineering scale of 6.4 × 10⁴ m³.

Eighteen thermal desorption engineering cases, listed in Appendix A, were selected as the comparison benchmark to verify the economic and energy consumption performance of the “gas ultra-high-temperature thermal desorption (UHT)–chemical oxidation (ISCO)” composite technology in this study. Although different heating methods, such as ERH and TCH, were included, the selection criteria were unified as “gas or electric heating ≥ 900 °C, scale ≥ 10,000 m³, and removal rate ≥ 99%”, and the technical boundaries were consistent with those of this study. Moreover, the cases covered 15 sites in North China, East China, South China, and Europe and America, with the lithology, pollutant types, and concentration ranges significantly overlapping with those of the local area.

The unit processing energy consumption was reduced to 820 kWh·m⁻³, which was 13.68% lower than the median energy ratio of 950 kWh·m⁻³ (Appendix A Table A1). The measured data confirm that the “high temperature–high speed–high heat recovery” pathway simultaneously satisfies reaction kinetics and maximizes system energy efficiency. This provides a replicable engineering paradigm for large-scale ex situ thermal desorption (ESTD) to overcome the “high temperature = high energy consumption” bottleneck.

4.3.2. Technology–Policy Synergies for Carbon Reduction Pathways

From a life cycle perspective, the natural gas-coupled waste heat recovery scheme exhibited a “fuel–system” dual carbon reduction effect. Its carbon factor (0.202 kg CO₂e·kWh⁻¹) was 22% lower than the coal-fired benchmark (0.258 kg CO₂e·kWh⁻¹), only 36% of the East China power grid average (0.557 kg CO₂e·kWh⁻¹). This advantage stems from three combined contributions: the low carbon content of natural gas, stepwise waste heat recovery (reducing the flue gas temperature from 500 °C to 200 °C), and high-temperature complete oxidation (reducing post-treatment energy consumption). Currently, natural gas remains the mainstream heat source for thermal desorption remediation. If a forward-looking switch to a zero-carbon heat source (70% photovoltaic-molten salt storage + 30% green electricity) is adopted, its carbon factor could be reduced to 0.035 kg CO₂e·kWh⁻¹, corresponding to an additional 8.8 kt CO₂e reduction [29,63].

Economic–policy coupling analysis shows that the current marginal abatement cost curve (MACC) is 16.71–25.07 USD t⁻¹ (estimated from 2023 engineering, procurement, and construction (EPC) quotations and annual emission reductions for resistance heat storage in East China), which is higher than the annual average price of the national carbon market (12.79 USD t⁻¹) [63]. However, including soil remediation in the “negative emissions” methodology after the 2025 industry expansion will fundamentally reverse this cost balance. Per 1 m³ of remediated soil, the carbon emission reduction potential (ΔC × E) was calculated as 0.167 kg CO₂e kWh⁻¹ (ΔC = 0.202–0.035) × 950 kWh m⁻³ (the mean of 18 cases; Appendix A Table A1) ≈ 0.159 t CO₂e m⁻³. Once the “soil remediation carbon credit” (SRCC) price reaches ≥20.89 USD t⁻¹, the zero-carbon heat source scheme will gain a net present value advantage.

Large and complex contaminated sites should prioritize the “high-parameter kiln + zero-carbon heat source” combination to provide decision-makers with clear signals across 2–3 carbon market cycles. This will allow them to seize the carbon credit premium window and achieve synergistic benefits from technology and policy alignment.

4.3.3. Cost–Benefit Mechanisms for Economies of Scale

When the processing scale exceeds the economic threshold of 5.2 × 10⁴ m³, the ESTD cost curve enters a “scale-induced gradual decline zone.” The total cost for the 7.5 × 10⁴ m³ project in this study was 111.42 USD m⁻³, 18.88% lower than the median energy ratio of 138.58 USD m⁻³ from 18 cases in the literature (Appendix A Table A1). The mean scale of these cases in the literature was 37,508.41 m³ (Appendix A Table A1), verifying the “scale–technology” synergistic cost reduction trend. While the 1000 °C high-temperature scheme required an upgrade in the grade of refractory materials and burners, the 33% reduction in residence time simultaneously increased the single-kiln processing capacity, resulting in a dominant fixed cost dilution effect [29]. Additionally, the high processing flow rate meant the marginal investment in “incremental technologies” (e.g., waste heat recovery and intelligent temperature control) was lower than their energy-saving benefits.

Inferential statistics were added to rigorously test the “better-than-benchmark” claim. A one-sample t-test against the literature mean of 138.58 USD m⁻³ gave t = −6.89 (df = 17; one-tailed p = 1.3 × 10⁻⁶), rejecting H₀ and confirming that the achieved unit cost (111.42 USD m⁻³) was significantly lower than the published benchmark.

The analysis of economies of scale is centered on direct costs: 52.9% for energy (electricity + natural gas), 13.6% for chemicals, 19.5% for labor and machinery, and 14.0% for temporary construction and environmental protection measures. Compared to values within the range of established experience (energy: 55–65%), the energy proportion is slightly lower, which verifies that the energy cost control effect of this project is good and can provide a reference for projects with the same process and scale.

This aligns with the observation in that “for scales > 5 × 10⁴ m³, energy consumption proportion decreases while integrated technology benefits become prominent.” Thus, economies of scale provide a “floor” for cost reduction, while technology integration further drives down costs. Together, they overcome the “diseconomies of scale” bottleneck of traditional thermal desorption, providing a replicable decision-making framework for large sites to select high-parameter kiln systems.

5. Conclusions

This study developed a “parallel diversion–high-temperature low-carbon” remediation paradigm for 7.5 × 10⁴ m³ of composite contaminated soil, designed based on differences in pollutant thermal stability and oxidizability. ESTD at 1000 °C combined with alkali-activated persulfate oxidation achieved a median removal rate of over 98% for PHC, PCP, B[a]A, D[ah]A, and B[a]P. More than 95% of the treated soil samples met the criteria of both the Chinese National Standard (GB) 36600-2018 and the United States Environmental Protection Agency (USEPA) regional screening level (RSL). By replacing the “one-size-fits-all” process with pollutant attribute-driven, precise routing, the project reduced the total combined cost by approximately 20.36% and unit energy consumption by approximately 13.68%. This breakthrough overcomes the traditional “diseconomies of scale” limitation associated with ESTD.

At the 10,000-ton engineering scale, quantification showed that when the remediation scale exceeded 5 × 10⁴ m³, the carbon intensity of the natural gas waste heat recovery scheme was 22% lower than that of the coal-fired benchmark. Life cycle scenario analysis further revealed that upgrading to a hybrid heat source (70% photovoltaic-molten salt storage + 30% green electricity) could reduce the carbon intensity to 0.035 kg CO₂e·kWh⁻¹. This provides a replicable technology–carbon reduction roadmap for subsequent high-temperature low-carbon remediation projects.

Engineering Implications and Future Outlook

This study verified a “scale–waste heat recovery” model for large and complex contaminated sites that enables a gradual transition to “high-parameter kilns + renewable energy heat sources,” thereby achieving net-zero remediation goals. The remediation industry urgently needs to shift from “end-of-pipe treatment expenditure” to “net-zero frontiers”, ensuring that every cubic meter of excavated soil contributes to both “zero pollution” and “net-zero emissions.”