Optimization of Dechlorination Conditions in the Co-Pyrolysis of Wood Sawdust and PVC: An Experiment

This article presents a systematic experimental study on optimizing the dechlorination conditions for the co-pyrolysis of wood sawdust and polyvinyl chloride (PVC), a key process for converting biomass-plastic mixed waste into low-chlorine oil, gas, and char products. It covers sample preparation, testing items, analytical methods based on international standards, single-factor and orthogonal optimization design, typical results, and industrial applications, serving as a practical technical reference for laboratories and engineers working on clean thermal conversion of chlorine-containing waste streams.

Technical Overview

Polyvinyl chloride is one of the most widely used plastics, containing approximately 56–57 wt% chlorine in its molecular chain. When PVC enters thermal conversion systems such as pyrolysis or incineration, the cleavage of C–Cl bonds releases hydrogen chloride (HCl), which causes severe corrosion of reactors and downstream equipment, poisons catalysts, catalyzes the formation of chlorinated aromatics such as chlorobenzene and chlorophenol in pyrolysis oil, and raises the risk of polychlorinated dibenzo-p-dioxin and furan formation. Direct pyrolysis of PVC therefore yields an oil product that is unsuitable for use as fuel or refinery feedstock without costly upgrading.

Co-pyrolysis of PVC with lignocellulosic biomass such as wood sawdust has emerged as an effective in-situ dechlorination strategy. Wood sawdust decomposes in a temperature window that overlaps the PVC dehydrochlorination stage (approximately 240–360 °C), and its intrinsic alkali and alkaline earth metals (K, Ca, Mg, Na), together with hydroxyl-rich char surfaces, chemically capture released HCl as stable inorganic chlorides retained in the solid residue. The optimization experiment described here quantifies the effects of pyrolysis temperature, sawdust-to-PVC blending ratio, heating rate, residence time, and carrier gas flow rate on chlorine distribution among char, oil, and gas, and identifies the operating window that minimizes the chlorine content of the liquid product while maintaining acceptable oil yield.

The core reactions involved include the dehydrochlorination of the PVC backbone (–CH2–CHCl– → –CH=CH– + HCl), the capture of HCl by biomass-derived metal oxides (for example, CaO + 2HCl → CaCl2 + H2O and K2O + 2HCl → 2KCl + H2O), and radical interactions in which hydrogen-donating fragments from cellulose, hemicellulose, and lignin suppress the incorporation of chlorine into organic molecules in the oil phase.

Testing Samples

Wood Sawdust

Pine or poplar sawdust collected from woodworking operations is air-dried, ground in a mill, and sieved to a particle size fraction of 0.25–0.85 mm (40–60 mesh). The sample is then oven-dried at 105 °C for 24 hours to a moisture content below 5 wt%. Characterization includes proximate analysis (moisture, volatile matter, ash, fixed carbon) by thermogravimetry in accordance with ASTM E1131, ultimate analysis (C, H, N, S; O by difference) following ASTM D5373 and ISO 16948, and calorific value determination per ISO 1928 or ASTM D5865. Typical pine sawdust contains 78–82 wt% volatile matter, 0.4–0.8 wt% ash, and measurable potassium and calcium that act as natural chlorine sorbents.

PVC Feedstock

The PVC sample may be virgin suspension-grade resin powder or washed and granulated waste PVC profile material. It is dried at 60 °C for 12 hours and sieved to the same particle size range as the sawdust to eliminate particle-size effects on heat and mass transfer. Total chlorine content is verified at approximately 56.5–57.0 wt%, and thermogravimetric analysis confirms the two characteristic mass-loss stages: dehydrochlorination at 240–360 °C and main-chain scission with polyaromatic formation at 400–520 °C.

Blended Samples

Sawdust and PVC are mechanically mixed at defined mass ratios, for example 10:0, 9:1, 4:1, 3:1, 1:1, 1:3, and 0:10, in batches of 5–10 g per run. Homogeneity is ensured by tumbling mixing for 30 minutes, and each blend is prepared in triplicate to assess feedstock variability. Blends are stored in sealed desiccators prior to testing to prevent moisture re-adsorption.

Testing Items

The optimization experiment evaluates the following response variables:

• Total chlorine content and chlorine distribution among char, oil, and gas phases, expressed as a percentage of chlorine in the feed.
• Dechlorination efficiency of the solid phase, defined as (chlorine in feed minus chlorine retained in char) divided by chlorine in feed, multiplied by 100%.
• Organic chlorine concentration in the pyrolysis oil, a direct indicator of oil quality and corrosivity.
• Product yields of char, oil (including aqueous phase), and non-condensable gas.
• HCl concentration in the gas phase captured in alkaline absorption solution.
• Higher heating value of the oil and char products.
• Chemical composition of the oil by GC-MS screening, with particular attention to chlorobenzene, chlorophenol, and chlorinated PAH species.
• Gas composition (H2, CH4, CO, CO2, C2–C3 hydrocarbons) by micro-gas chromatography.
• Morphology, surface chemistry, and elemental composition of the char by FTIR, SEM-EDS, and XRF.

Testing Methods

Thermogravimetric Analysis with Evolved Gas Detection

TGA and TGA-FTIR coupling are used to map the decomposition behavior of the individual feedstocks and their blends. Runs are performed under nitrogen at 50–80 mL/min with heating rates of 5, 10, 20, and 40 °C/min from ambient to 800 °C. Overlapping of the sawdust cellulose/hemicellulose decomposition peaks with the PVC dehydrochlorination peak indicates the temperature window in which biomass-derived sorbents are most active, providing the basis for selecting factor levels in the pyrolysis experiments.

Fixed-Bed Pyrolysis Experiments

Pyrolysis is conducted in a quartz tubular reactor (inner diameter 40–60 mm) mounted in a programmable electric furnace. A weighed sample boat holding 5–10 g of material is inserted into the pre-set isothermal zone, and high-purity nitrogen at 100–400 mL/min purges the system. Test temperatures span 300–700 °C, heating rates 5–40 °C/min, and holding times 15–60 min. Volatiles pass through an ice-water condensation train held at −5 to 0 °C for oil recovery, followed by three serial absorption bottles filled with 0.1 mol/L NaOH solution to quantitatively trap HCl. Residual non-condensable gas is collected in gas bags for analysis. Each condition is tested in triplicate to establish error bars.

Chlorine Determination

Chlorine in the solid feed and char is determined by the Eschka mixture combustion method followed by potentiometric titration or ion chromatography, consistent with ISO 587 and GB/T 3558 for solid fuels. Chlorine in the pyrolysis oil is measured by oxygen bomb or oxygen flask combustion with subsequent anion determination by ion chromatography, following the principles of ASTM D808 and ASTM D4327. Chlorine in solid recovered fuels may additionally be cross-checked against EN 15408, and water-soluble chloride in biomass against ISO 16995. HCl in the absorbed gas fraction is quantified directly by ion chromatography. A chlorine mass balance closure of 95–105% is required for a run to be accepted.

Experimental Design and Statistical Optimization

The optimization follows a two-step strategy. First, single-factor experiments identify the individual effect and plausible range of each parameter. Second, an orthogonal array (for example L16 with four factors at four levels) or a Box-Behnken response surface design is applied, with pyrolysis temperature, sawdust-to-PVC ratio, heating rate, and residence time as factors, and oil chlorine content and dechlorination efficiency as primary responses. Analysis of variance (ANOVA) identifies the significant factors and their interactions, and a verification experiment under predicted optimal conditions confirms the model with a relative standard deviation below 5% across three replicate runs.

Optimization Results and Discussion

Effect of Pyrolysis Temperature

Temperature is consistently the most significant factor. Between 300 and 500 °C, raising the temperature accelerates dehydrochlorination and increases the release of chlorine from the solid phase, with dechlorination efficiency typically exceeding 90% above 450 °C. However, above approximately 550–600 °C, secondary reactions promote the conversion of HCl into thermally stable chlorinated aromatics in the oil, so the organic chlorine in the liquid product passes through a minimum. The optimal temperature for oil quality is therefore generally found in the window of 480–550 °C.

Effect of Sawdust-to-PVC Blending Ratio

Increasing the biomass proportion in the blend reduces oil chlorine content sharply at first and then gradually levels off. At sawdust-to-PVC ratios near 4:1, the alkali metal and hydroxyl population is typically sufficient to capture the majority of released HCl, lowering oil chlorine from values above 1 wt% for PVC-rich blends to below 0.1 wt%. Excessive dilution beyond roughly 9:1 yields diminishing dechlorination benefit while reducing the plastic-derived oil fraction, so ratios between 3:1 and 5:1 represent the practical optimum for most feedstocks.

Effect of Heating Rate, Residence Time, and Carrier Gas Flow

• Moderate heating rates of 10–20 °C/min allow dehydrochlorination to approach completion before extensive tar formation, favoring chlorine capture in the char matrix; very fast heating favors organic chlorine formation in the oil.
• Residence times of 30 minutes at the final temperature are sufficient for complete devolatilization; longer holds increase energy consumption without further chlorine reduction.
• Nitrogen flow rates of 200–300 mL/min effectively sweep HCl out of the reaction zone and reduce secondary chlorination of oil vapors; excessive flow dilutes condensable vapors and reduces oil yield.

Two-Stage Pyrolysis Strategy

A particularly effective process intensification is temperature-programmed two-stage pyrolysis: the blend is first held at 310–330 °C for 20–30 minutes, corresponding to the peak HCl release window, and then ramped to 500–520 °C for oil production. In the first stage, the majority of chlorine exits as HCl while biomass char with active sorption sites is formed; in the second stage, the newly generated oil vapors contact a chlorine-depleted, sorbent-rich char bed. Reported oil chlorine levels under this strategy can fall below 0.05 wt%, approaching the specification limits required for co-processing in conventional refineries.

Verified Optimal Conditions

Representative optimized conditions from orthogonal verification tests are: pyrolysis temperature 500 °C, sawdust-to-PVC mass ratio 4:1, heating rate 15 °C/min, holding time 30 minutes, and nitrogen flow 250 mL/min. Under these conditions, typical results include a solid-phase dechlorination efficiency above 95%, chlorine distributed as roughly 50–65% HCl in gas (recovered from the absorption train), 20–30% inorganic chloride fixed in char, and less than 10% in the oil, with oil chlorine reduced to the 0.03–0.10 wt% range and oil higher heating value in the order of 25–33 MJ/kg depending on blend composition.

Applications

• Co-processing of forestry residues and PVC-containing waste plastics from furniture manufacturing, construction and demolition streams, and cable sheathing.
• Production of low-chlorine pyrolysis oil suitable as industrial furnace fuel or as a co-feed for fluid catalytic cracking and hydrogenation units after polishing.
• Pre-dechlorination of solid recovered fuel (SRF/RDF) blends to reduce boiler corrosion and acid gas emission control costs in cement kilns and waste-to-energy plants.
• Recovery of concentrated HCl from the gas absorption stream as a saleable by-product for steel pickling and chemical synthesis.
• Utilization of chlorine-fixed biochar as a solid fuel, adsorbent material, or soil amendment after leachability assessment.
• Design basis for continuous screw or rotary-kiln co-pyrolysis pilot plants treating mixed biomass-plastic feedstocks.

FAQ

Why does co-pyrolysis with wood sawdust reduce chlorine in the oil?

Wood sawdust contributes alkali and alkaline earth metals (K, Ca, Mg, Na) and oxygen-rich functional groups that react with HCl released from PVC and fix it as stable inorganic chlorides in the char. In addition, hydrogen-donating radicals from biomass decomposition stabilize chlorine-containing intermediates, suppressing the formation of chlorinated organic compounds in the oil phase.

What is the most important operating parameter?

Pyrolysis temperature, followed closely by the sawdust-to-PVC blending ratio. ANOVA of orthogonal experiments typically shows these two factors, together with their interaction, account for the dominant share of the variance in oil chlorine content and dechlorination efficiency.

How is chlorine in the products measured?

Solid samples are combusted with Eschka mixture and the chloride is quantified by titration or ion chromatography per ISO 587 and GB/T 3558; oil samples undergo oxygen bomb or flask combustion with ion chromatographic finish per ASTM D808 and ASTM D4327 principles; gas-phase HCl is trapped in NaOH solution and measured by ion chromatography. Mass balance closure above 95% validates each run.

What chlorine level in the oil is acceptable?

For direct combustion as furnace fuel, oil chlorine below approximately 0.1 wt% is commonly targeted to limit corrosion. For co-processing in refineries, specifications are far stricter, typically in the tens of ppm range, which is why two-stage pyrolysis or post-treatment upgrading is recommended.

Can the process be scaled up?

Yes. The optimized parameters translate directly to continuous screw-conveyor and rotary-kiln reactors, provided that staged heating, adequate vapor residence time control, and HCl-resistant materials such as quartz-lined or alloy-316L-plus-coating construction are incorporated, and the acid gas scrubbing train is sized for the full chlorine load.