Industrial Ventilation and Exhaust Hoods: Method

Hood Capture, Duct Balancing, Fan Selection and Dilution Ventilation

What the tool does

A local exhaust system is designed from the hood outward: the hood sets the exhaust flow needed to capture the contaminant, the branch duct carries it at a velocity that keeps particles moving, the main collects the branches, and the air cleaner, fan and stack follow. The tool walks that path. For each hood it computes the exhaust flow from the hood type, sizes the branch to the transport velocity, finds the hood static pressure and the losses of the branch, balances the branches where they meet the main (by design or with blast gates), sizes and loses the main segments, adds the air cleaner, finds the fan total and static pressure and power, and checks the stack exit velocity and height. A second part is dilution ventilation for a contaminant released into the room air; a third balances the makeup air and its heating load. The default case is an illustrative two-hood system (a flanged slot over a bench and a canopy over a hot process) and the xylene dilution case of MIL-HDBK-1003/17C Appendix B; every relation is checked by the test script against the printed examples of MIL-HDBK-1003/17C, the OSHA Technical Manual and 29 CFR 1910.94, and against an independent numpy / scipy script.

Hood exhaust flow

The published capture relations (NIOSH 76-162 hood figures; the open-hood forms are the Dalla Valle relations, which the ACGIH manual also prints) give the flow that produces the capture velocity V_c at the farthest point of release, a distance X from the hood face:

• plain opening, freely suspended: Q = V_c (10 X² + A), with A the face area;
• flanged opening, or an opening on a bench or floor (the surface acts as a flange): Q = 0.75 V_c (10 X² + A);
• free slot of length L: Q = 3.7 L V_c X; flanged slot: Q = 2.6 L V_c X (NIOSH prints 2.8 for a slot on a table, 2.6 with a wide flange; the formulae hold for X/L below about 0.3, beyond which the opening forms apply);
• canopy open on four sides, perimeter P, at a height H above the source: Q = 1.4 P H V_c; a canopy with sides closed uses the open perimeter in place of 1.4 P;
• booth, enclosure or downdraft grille: Q = V_f A_f, the face velocity times the open area.

The capture velocity itself is a judgement about the process: the public guidance gives 50 to 100 ft/min for vapours released with little energy into quiet air, 100 to 200 ft/min for active generation into moving air, 200 to 500 ft/min for high-energy release (grinding, blasting), and 500 to 2,000 ft/min for the most violent operations; OSHA's 1910.94(b) requires at least 200 ft/min into a portable-grinding enclosure, MIL-HDBK-1003/17C Table 2 gives 50 to 150 ft/min for open-surface tanks by hazard class, and NIOSH 76-162 specifies 100 to 200 ft/min for most machining and foundry hoods. The ACGIH manual's capture-velocity table, which refines these ranges, is cited and not reproduced. The OSHA Technical Manual's rule of thumb is that the source should sit within 1.5 duct diameters of a simple hood, where the capture velocity is roughly a tenth of the duct velocity, and that it falls another ten-fold at two diameters.

Velocity pressure, entry loss and hood static pressure

Velocity pressure is the kinetic pressure of the moving air, VP = ρ V² / 2. With standard air at 0.075 lb/ft³ and pressures in inches of water this is the handbook's VP = (V / 4,005)² (the exact constant at the dictionary's 249.082 Pa per inch of water is 4,008); the tool uses the density entered, so a hot or high-altitude system is handled by its density factor as in MIL-HDBK-1003/17C Appendix D. Air entering a hood loses h_e = F_h VP_d at the duct entry; a slot hood loses F_s VP_s at the slot as well. Published factors: 0.25 for a tapered takeoff with a 90° included angle (MIL-HDBK-1003/17C Appendix A), 0.45 tapered and 0.65 straight for the OSHA grinder hoods of 1910.94(b), 0.49 for a flanged round duct end (NIOSH 76-162 foundry figures), 1.78 slot velocity pressures for the slot of a plating-tank hood (MIL-HDBK-1003/17C), 1.18 in the NIOSH figures, 1.0 slot VP plus 0.5 branch VP for the vertical-spindle grinder of 1910.94 Figure G-1. NIOSH's Figure 1 gives the tapered-hood factor against the included angle, from about 0.9 for an abrupt entry to its minimum near 45°; the ACGIH entry-loss figures are cited only. The hood static pressure is SP_h = h_e + VP_d, the suction a gauge reads a few diameters down the duct, and the coefficient of entry C_e = √(VP_d / SP_h) = √(1 / (1 + F_h)) is the ratio of the actual flow to the ideal (OSHA Technical Manual III:3). A measured SP_h and the known C_e give the flow in service: Q = 4,005 C_e A √SP_h in US units.

Transport velocity and duct sizing

Dust-carrying ducts must keep the particles moving. The public minimums: 1,000 to 2,000 ft/min for vapours, gases and fumes, 2,500 ft/min for the mist of plating tanks (MIL-HDBK-1003/17C Appendix A), 3,500 ft/min in the main and 4,500 ft/min in the branch for grinding (29 CFR 1910.94(b)), 3,500 to 4,500 ft/min for foundry dust and 4,000 ft/min for woodworking (NIOSH 76-162, MIL-HDBK-1003/17C 10.4.3), 5,000 ft/min for abrasive blasting (UFC 3-410-04). The duct area is Q / V_t and the diameter is rounded down to a standard size so the velocity stays above the minimum, as the MIL-HDBK design examples do. Friction is Darcy's f (L / D) VP with the Colebrook friction factor at the Reynolds number and relative roughness, the same routine as the Pipe Flow tool; the ASHRAE friction chart and the equal-friction method present the same physics graphically. Fittings lose ΣC × VP: a smooth 90° elbow of r/D 1.5 about 0.15 to 0.2 (five-gore 0.11), a 45° elbow 0.06, a 45° branch entry about 0.28, a mitered elbow without vanes 1.2, an abrupt duct end 0.5 to 0.9; the values are entered per branch and segment.

Balancing at the junctions

Two paths meeting at a junction must arrive at the same static pressure, or the air will redistribute itself until they do. The velocity pressure method (balance by design) sizes the system so that this holds at the design flows: at each junction the static pressure of the main (hood 1 through the upstream segments) is compared with that of the entering branch; the larger governs. If the ratio is within about 1.05 nothing is done; between 1.05 and 1.20 the lower side's flow is raised to Q_corr = Q √(SP_governing / SP_lower), since every loss in the branch varies as the square of the flow (the friction factor's slow change with Reynolds number is recomputed, so the corrected branch lands within a fraction of a percent of the governing pressure); above 1.20 the handbook practice is to redesign the lower branch with a smaller duct or more loss rather than carry the extra air, and the tool flags the ratio. When the main is the lower side the whole upstream subsystem is raised together, through the duct sizes already chosen. A slot hood's slot area is set by the design flow at the entered slot velocity, so a corrected flow raises the slot velocity in that area and the slot loss scales with the flow like every other loss. Where the merged flow's velocity pressure in the next segment exceeds the governing duct's velocity pressure, the difference is added as an acceleration loss (the resultant-VP rule). The blast-gate method keeps the design flows and reports, at each junction, the pressure the gate in the lower branch must throttle; UFC 3-410-04 asks for gates to be locked and marked after balancing and never used to cut the airflow of a hood. Both methods give the fan the governing pressures.

Fan and stack

The fan sees the suction at its inlet (the main's static pressure plus the air cleaner) and must push the air through the outlet duct and stack to the atmosphere, where the exit velocity pressure is lost (an evasé recovers part of it; a weather cap adds loss and spoils the plume, which is why MIL-HDBK-1003/17C and the OTM reject caps). Fan total pressure is the rise in total pressure, FTP = (SP_out + VP_out) − (SP_in + VP_in); fan static pressure, the figure fan catalogues are rated on, is FSP = FTP − VP_out. Air power is Q × FTP (cfm × in. w.g. / 6,356 hp); the brake horsepower is the air horsepower divided by the fan's mechanical efficiency (OTM glossary) and the motor input adds its own efficiency. Select the fan at the total flow and this FSP on its curve (the Fan Operating Point tool), with the 6-and-3 rule of straight duct at inlet and outlet so the catalogue curve applies; a fan on hot or thin air develops pressure in proportion to density. For the stack, MIL-HDBK-1003/17C asks for an exit velocity of at least 3,000 ft/min (15.2 m/s) so rain and condensation do not run down the stack and for a discharge-to-wind velocity ratio of at least 1.5 (the OTM says 1.4) so the plume breaks away; it places the discharge at 1.5 times the building height (its 4.3.5), above the building's eddy zone, and the OTM asks for 10 ft above any roof line or intake within 50 ft, the stack downwind of intakes and at least 50 ft from them.

Dilution ventilation

General exhaust lets the contaminant into the room and dilutes it. A contaminant generated at the vapour rate G in a room exhausted at Q with a mixing factor K reaches C = K G / Q at steady state; the OSHA Technical Manual writes Q_d = q_d K_m 10⁶ / C_a with C_a in ppm, K_m from 2 to 5 (2 is optimum) and C_a typically half the PEL. The vapour rate of an evaporating liquid is its molar rate times the molar volume, 387 ft³ per lb-mol at 70 °F; MIL-HDBK-1003/17C Appendix B writes the same as Q = 403 × 10⁶ SG ER / (MW TLV) with the evaporation rate in pints per minute (403 = 387 × 8.33 / 8), and its xylene case (2 pints per hour, SG 0.86, MW 106.2, 100 ppm) gives 1,088 cfm with perfect mixing. The transient follows the mass balance V dC/dt = G − (Q/K) C, so C(t) = C_ss + (C₀ − C_ss) e^−t/τ with the time constant τ = K V / Q; after the generation stops the room purges from C₁ to C₂ in τ ln(C₁ / C₂). For fire and explosion the flow keeps the mixture at a fraction 1/SF of the lower explosive limit: Q = G SF 100 / (LEL B), with SF = 4 for 25 % of the LEL and B a temperature factor the handbook sets at 1 above 250 °F and 0.7 below (its own 85 °F example uses 1; the tool takes B as entered). 29 CFR 1910.94(c)(6)(ii) states the same 25 % rule per gallon of solvent sprayed, 4 (100 − LEL) × (ft³ of vapour per gallon) / LEL, 8,564 ft³ per gallon of toluene. Sensible heat is carried off at Q = q / (ρ c_p ΔT), the familiar cfm = Btu/h / (1.08 ΔT °F) at standard air. The OTM reserves dilution for low-hazard vapours released uniformly; MIL-HDBK-1003/17C notes ACGIH's advice against it for substances with a TLV below 100 ppm. Exposure limits: the tool offers the OSHA permissible exposure limits of 29 CFR 1910.1000 Table Z-1 (toluene from Table Z-2) as fills; ACGIH threshold limit values are cited, not reproduced.

Makeup air

Every cubic foot exhausted must come in from somewhere; without a replacement air system the room goes negative, infiltration whistles through cracks and doors, and the exhaust flow drops as the fan climbs its curve (OTM). MIL-HDBK-1003/17C modulates the replacement air, not the exhaust, to hold the room at its design pressure (−0.02 to −0.06 in. w.g. in dirty spaces), and distributes it so it does not disturb the hoods. The makeup required is the total exhaust less any other supply, and its winter heating load is ρ c_p Q ΔT, 1.08 × cfm × ΔT °F Btu/h at standard air; the Room Heating and Cooling Load tool carries the rest of the space load.

Sources and limits

Hood relations and loss factors: NIOSH 76-162 (1976) and the OSHA Technical Manual Section III Chapter 3; worked design arithmetic: MIL-HDBK-1003/17C Appendices A, B and D; duct design, blast gates, stacks and replacement air: UFC 3-410-04 (2017) and MIL-HDBK-1003/17C; exposure limits and the LEL rule: 29 CFR 1910.1000 and 1910.94. The ACGIH Industrial Ventilation manual and the ASHRAE Handbook are cited only; their capture-velocity, entry-loss and fitting tables are entered values here. Not covered: open-surface tank exhaust rates by tank aspect ratio (MIL-HDBK-1003/17C Tables 3 and 4), push-pull hoods, low-volume high-velocity tool hoods, rectangular ducts, system-effect factors at the fan, duct leakage, and the dispersion modelling that a stack near taller buildings needs.