Industrial air pollution from solid particulates is one of the primary compliance challenges in process engineering. Processes such as cement grinding, coal combustion, and mineral processing generate dust loadings that can exceed regulatory emission limits, accelerate equipment wear, and create serious occupational health hazards [1]. Two well-established control technologies — cyclone separators and fabric bag filters — are widely deployed, often in series, to address this problem.

Cyclone separators use centrifugal force to separate coarse and medium particles from a gas stream without consumables or moving parts. Their advantages include low capital cost, continuous dry operation, and tolerance of high temperatures and pressures [2]. However, cyclone collection efficiency falls sharply for particles below 10 μm due to the low inertia of fine particles, which tend to follow gas streamlines rather than migrate to the cyclone wall [3].

A downstream bag filter compensates for this weakness by capturing fine particles through interception, inertial impaction, diffusion, and dust-cake filtration mechanisms [4].

This paper presents the design of a combined cyclone–filtration system modelled in DWSIM, with performance evaluated across particle sizes from 1 μm to 50 μm. The specific objectives are: (i) to develop a process model of the cyclone–filtration system in DWSIM; (ii) to evaluate particle removal efficiency as a function of particle size; (iii) to analyse pressure drop under selected operating conditions; and (iv) to compare the performance of the cyclone alone against the integrated system. The novelty lies in applying DWSIM — an open-source, equation-oriented process simulator — to systematically model both stages of a particulate-control train and quantify the benefit of hybrid operation.

Zabala-Quintero et al. [5] used CFD to characterise cyclone efficiency and pressure drop, demonstrating that separation performance is strongly dependent on inlet velocity and cyclone geometry. Feng et al. [3] investigated fine-particle separation under varying operating conditions, establishing the trade-off between cut diameter and pressure drop. Teng et al. [4] developed and experimentally validated a model for filter pressure drop during dust-cake loading of pleated filters.

Sylvia et al. [6] simulated a cyclone integrated with a bottom-ash bed filter for PM2.5 removal in a palm oil mill, achieving combined efficiencies exceeding 95%. A consistent finding across these studies is that neither cyclone-only nor filter-only configurations provide optimal solutions. The research gap addressed here is the absence of a readily reproducible, simulation-based design study that quantifies the improvement from cyclone–filtration integration across the full sub-50 μm particle range and explicitly evaluates the pressure-drop cost of that improvement.

A. Process Configuration

The process consists of four elements in series: (1) a dust-laden gas inlet stream; (2) a tangential-entry cyclone separator; (3) a pulse-jet bag filter; and (4) a clean-gas outlet. The cyclone removes coarse and medium particles by centrifugal action; the bag filter captures residual fines. The DWSIM flowsheet was constructed using material stream blocks, a gas–solid separator unit, and a filter unit, with mass balances enforced at each node.

B. Cyclone Design Equations

The Lapple model [1] was used to estimate the cyclone cut diameter d_pc, the particle size at 50% collection efficiency:

where μ is gas dynamic viscosity (Pa·s), b is cyclone inlet width (m), N_e is the effective number of gas turns, ρ_p is particle density (kg/m³), and V_t is tangential velocity (m/s). Grade efficiency was computed from:

Cyclone pressure drop was estimated as:

where N_H is the number of velocity heads (dimensionless).

C. Filter Design Equations

Clean-filter pressure drop was estimated using Darcy’s law [4]:

where L is filter medium thickness (m), V_f is face velocity (m/s), and k is permeability (m²). Combined system efficiency was calculated as:

D. Design Basis

Table I summarises the input values used throughout. Gas was modelled as ambient air (ρ = 1.2 kg/m³, μ = 1.8 × 10⁻⁵ Pa·s). Particle density was set at 1500 kg/m³, representative of cement or mineral dust. The reference inlet velocity of 18 m/s is within the standard operating range for industrial cyclones [1]. Particle sizes from 1 μm to 50 μm were evaluated to span the range from PM2.5-range fines to coarse industrial dust.

A. Particle Removal Efficiency

Table II presents collection efficiency for the cyclone alone, the filter alone, and the combined system. The cyclone efficiency follows the expected Lapple grade-efficiency relationship: 3.8% at 1 μm, rising steeply through 50.0% at 5 μm (near the cut diameter of 4.37 μm), to 99.0% at 50 μm. This confirms that the cyclone is effective for particles well above its cut diameter but unreliable for fine particulate matter.

The integrated system dramatically improved fine-particle capture. At 1 μm, combined efficiency reached 88.5%, driven almost entirely by filtration since the cyclone contributed only 3.8%. At 5 μm and above, combined efficiency was ≥98.0%, meeting or exceeding the requirements of many national ambient air quality standards for PM10 and PM2.5 control [1].

B. Residual Dust Loading

Based on an inlet dust concentration of 100 mg/m³, the cyclone alone produced an outlet loading of 96.2 mg/m³ for 1 μm particles, unacceptable for most emission limits. The combined system reduced this to 11.5 mg/m³ — an 84.7 percentage-point improvement. For 10 μm particles, the cyclone alone left 20.0 mg/m³ while the integrated system achieved 0.4 mg/m³. At 20 μm and above, both systems performed well, though the integrated system consistently returned near-zero residual loadings.

C. Pressure Drop Analysis

Table III summarises the pressure drop results at the reference condition. The cyclone accounted for 1166.4 Pa (95.6%) of the total 1220.4 Pa clean-system pressure drop. The clean-filter drop was only 54.0 Pa, confirming that the filter does not significantly increase system resistance when the filter medium is clean. In practice, dust-cake accumulation will raise filter resistance progressively; periodic pulse-jet cleaning is required to maintain acceptable operating pressure drop [4].

D. Velocity Sensitivity

Increasing inlet velocity from 12 to 24 m/s reduced the Lapple cut diameter from 5.35 μm to 3.78 μm, improving fine-particle capture. However, cyclone pressure drop rose from 518.4 Pa to 2073.6 Pa — a fourfold increase. This non-linear relationship (ΔP ∝ V²) means that chasing fine-particle separation purely through velocity escalation rapidly becomes energy-inefficient. The reference velocity of 18 m/s was chosen as the optimum, balancing a cut diameter of 4.37 μm against a pressure drop of 1166.4 Pa, aligning with guidance from US EPA [1] and the findings of Feng et al. [3].

The global industrial dust collection equipment market was valued at approximately USD 6.4 billion in 2023 and is projected to grow at a CAGR of 5.1% through 2030, driven by tightening emissions legislation in Asia-Pacific, the Middle East, and sub-Saharan Africa [7]. The proposed hybrid cyclone–filtration system targets cement plants, quarries, and chemical processors in the Gulf Cooperation Council (GCC) region, where the regulatory environment is tightening in line with WHO ambient air quality guidelines.

The competitive advantage rests on three pillars. First, DWSIM-based pre-engineering reduces design risk by allowing performance to be validated computationally before capital is committed. Second, the two-stage configuration reduces filter bag replacement frequency by removing coarse particles in the cyclone stage — a direct operating cost saving. Third, standardised modular design enables replication across multiple plant units with minimal re-engineering. Revenue streams include equipment supply, commissioning, maintenance contracts, and filter bag replacement subscriptions. The primary market entry route is through EPC contractors serving refinery and cement projects in Bahrain, Saudi Arabia, and the UAE.

This paper demonstrated that a combined cyclone–filtration system, designed and simulated in DWSIM using the Lapple model and Darcy’s law, significantly outperforms a cyclone-only system for industrial particulate control. Key findings: (i) the cyclone cut diameter at 18 m/s is 4.37 μm; (ii) combined system efficiency exceeds 88% at 1 μm and 98% at 5 μm and above; (iii) total clean-system pressure drop is 1220.4 Pa, dominated by the cyclone stage; and (iv) increasing inlet velocity improves fine-particle capture but raises energy consumption quadratically.

The principal limitation is that results are based on idealised steady-state assumptions — constant gas properties, mono-disperse particle classes, and clean-filter conditions — without experimental validation. Recommended future work includes: laboratory-scale validation of the DWSIM model; transient modelling of dust-cake growth and filter pressure drop evolution; use of actual industrial particle-size distributions; and a full life-cycle cost analysis comparing this hybrid system against electrostatic precipitators and venturi scrubbers.