Exterior Hoods
10.2.2.1 Design Considerations Genera/
An exterior hood is often a natural choice for an exhaust. It is usually easy to install, less expensive, and does not need any large changes in the outlay of the room or the process. Often it is possible to connect this type of hood to an existing exhaust duct system and when the flow rate is relatively small, the existing supply air system may be maintained without changes.
|
V
► Contaminant flow
|
FIGURE 10.6 Definition of capture efficiency, A. M = contaminant source rate, m = contaminant transport (directly) into the exhaust hood, A = m/M.
As mentioned before, the efficiency of an exterior hood may be low, which is a disadvantage of this type of system. Before an exterior hood is chosen other alternatives should be investigated regarding efficiency, suitability, and cost.
Exterior hoods are most favorable when one or more of the following conditions are fulfilled:
The exhaust must be very close to the source and it is not possible to surround the source by an enclosure;
The generation has one principal direction and it is necessary to have a distance between the source and the exhaust;
The source is or could be movable; and
The source is temporary in place or time.
The first of these conditions is the most important factor when deciding to use an exterior hood. Exterior hoods are allowed when the demands on the exhaust, and the hazard of the contaminants, are moderate.
The exterior hoods described here are divided into basic openings, rim exhausts, low-volume high-velocity (LVHV) hoods, receptor hoods (canopy hoods), and downdraft ventilation tables. Many varieties of these types of hoods exist. Some of these have been described and investigated more thoroughly than others because they are used more often or they are of more general use and applicability than the more specialized hoods.
The design equations and parameters for the different hoods are quite different, which is the main reason for describing each of these hood types separately.
II |
|||
Fume cupboard |
|||
4 |
C O Measurement |
||
Point |
|||
N C |
M-— |
||
Opening for work and airflow |
|||
Tracer gas source |
|||
/ |
|||
• ■ ■ Tracer gas flow |
FIGURE 10.7 Definition of protection factor, pF (This is one definition of containment efficiency; see j. Pekkinen et al.1) G = gas tracer flow rate, mg min-1, C = concentration in breathing zone, mg rrr J. PF = logio (G/QLocation of Supply and Exhaust Openings and Sources
The supply air for exterior hoods should be brought into the room without disturbing the exhaust flow. The supply openings are usually placed far away from the exhausts and it is necessary to check that the supply air velocity near the exhaust is much lower than the velocity due to the exhaust. If the supply air velocities are less than one-tenth of the exhaust air velocities at distances less than three exhaust opening diameters from the exhaust, then the influence of the supply on the exhaust should be negligible. This is also true for cross-drafts caused by other activities. For cases with higher supply air velocities or higher cross-draft velocities, supply air velocity or cross-draft velocity must be included in the design of the exhaust, the supply air or cross-draft should be redirected, or the type of exhaust should be changed to one with less sensitivity to supply air or cross-draft velocities.
The exhaust opening should be placed as close as possible to the source. This normally means a distance less than one exhaust opening diameter between source and exhaust opening. If the distance is larger both opening and flow rate must be increased so much that it may be infeasible to use an exterior hood. For sources with one principal generation direction, the distance could be larger with only slightly less efficiency. One example is a canopy hood over a hot bath.
The exhaust could be placed above, below, or beside the source depending on generation direction. If possible, the main exhaust direction (center axis of opening area) should be the same as the main contaminant generation direction. Small changes in location, both sideways and in distance, could result in large changes in the efficiency of exterior hoods. If the exhaust is designed to have the highest efficiency in one specific placing, small moves could decrease the efficiency substantially.
For the supply opening in relation to the source, the same arguments are valid as for the supply in relation to the exhaust. However, it is sometimes possible to use the supply air to increase the transport from the source into the exhaust; these types of systems are described in combined exhaust and supply systems (Section 10.4).
Location of Workers and Tools Relative to Exhaust and Source
The worker and the tool should be outside the exhaust system’s distance of influence and the source should be inside this distance, as is shown principally in Fig. 10.8.
This is usually not a problem with small exhausts. With large exhausts (opening area larger than 0.1 m2), parts of a worker or of tools could easily come between the source and the exhaust. This diminishes the efficiency and could also spread the contaminant in unwanted directions, even if the exhaust velocity is
|
|
Exhaust airflow |
Exhaust hood |
W
|
Airflow —’► Contaminant flow
—————— Capturing limit for exhaust
M. Contaminant source W, Worker T, Tool
■■ FIGURE 10.8 Schematic ideal placing of worker, tool, and contaminant source when using exterior hoods.
Large enough to transport the contaminant into an undisturbed exhaust. With still larger exhausts, a worker too close to the source could be exposed to the contaminant due to vortices created when air flows around the body. See Sections 7.3 and
10.3.4 on airflow around obstacles.
1’ools are normally close to the opening. One way to diminish the disturbances from a tool is to shape the opening around that part of the tool when — contaminants are generated. Another is to size the opening specifically for the source and place it very close to the source and away from disturbances from the tool.
Since exterior hoods are unshielded, nearly all disturbances can dramatically change the performance of the hood. Changes in direction of contaminant generation could result in contaminant spread outside the exhaust; a variable generation rate could result in intermittent contaminant spread directly to the surroundings when the source volume rate becomes larger than the exhaust flow rate; tools moving around or rotating could either disturb the flow field or redirect the contaminants or both; cross-drafts from moving persons or vehicles, or leakage through door openings or wall cracks could temporarily or permanently disturb the exhaust flow field and result in spreading of the contaminant in the room.2’3 A moving source demands a moving exhaust or a very large opening to ensure that the hood is as close as possible to the source. An exhaust can be moved virtually, i. e., the openings are not moved, but instead the exhaust flow rate is connected to different openings or parts of one opening (at different locations) depending on where the moving source is at that moment. This is a way of minimizing the exhaust flow rate without diminishing the efficiency. This type of hood puts high demands on the control system for the exhaust. The misplacing of a baffle in the exhaust opening (to diminish the necessary flow rate) could also result in contaminants not being captured.
A laboratory study of surface-treatment tanks by Braconnier et al.4 showed the effects of cross-drafts and obstructions to airflow on capture efficiency. They found that, without obstructions, capture efficiency decreased with increasing cross-draft velocity but the importance of this effect depended on freeboard height. In their study, cross-draft direction was always perpendicular to the hood face and directed opposite to the hood suction flow. For low cross-draft velocities (less than 0.2 m s-1), efficiency remained close to 1.0 for the three freeboard heights studied. With higher cross-draft velocities, efficiency decreased as freeboard height decreased. For example, when the crossdraft velocity was 0.55 m s’1, efficiency decreased from 0.90 to 0.86 to 0.67 as freeboard height decreased from 0.3 m to 0.1.5 m to O.’l m, respectively.
A similar effect was observed for changes in hood flow rate. With a fixed cross-draft velocity, capture efficiency decreased with decreasing hood flow rate. This effect was much more important when freeboard height was small. Their results showed that when hood flow rate was 1.5 m’s-1 nr2, efficiency remained close to 1.0 as long as the cross-draft velocity was less than 0.45 in s-1. The most severe conditions tested were a hood flow rate equal to 0.33 m3 s_l nr2 and crossdraft velocity equal to 1.15 m s_1. Under these conditions, capture efficiency was equal to 0.83 for freeboard height equal to 0.3 m, but decreasing to 0.4 when freeboard height was decreased to 0.1 m.
The effects of obstacles in the flow field were also studied. A bar with dimensions of 0.04 m high x 0.003 m wide x 1 m long was used. Capture efficiency was measured with the obstacle at various heights along the center of the tank (parallel to the exhaust hood). For the experimental conditions studied (hood flow = 0.33 m-? sr1 nr2; cross-draft velocity = 0.5 m s-1; freeboard height = 0.3 m), efficiency was reduced from 0.8 when no obstacle was present to 0.28 when the obstacle was removed from the tank to a height of 0.1 m above the rim of the tank. When the obstacle reached a height of 0.4 m, its influence was no longer noticeable. Moving the obstacle upstream of the tank also resulted in effects on hood capture. When the obstacle was placed just above the height of the slot hood, higher capture efficiency resulted. When the obstacle was moved higher in the flow field, efficiency decreased, especially for low hood flow rates. Smoke tests showed the location where the change takes place to be approximately at the height where the flow separates between flow entering the slot hood and flow continuing downstream of the tank. This is similar to the dividing streamline described by Conroy and Ellcnbecker — ‘ and Flynn and Ellenbecker67 and to capture envelopes determined experimentally by Fletcher and Johnson.8
A field study of vapor degreasing tank emissions arid hood capture efficiency was performed by Conroy et al.9 Solvent emissions to the local ventilation system and to the workspace were measured at 13 sites under operating conditions. In this study, no correlation w’as found between local ventilation capture efficiency and hood flow rate. The authors found that the ACGIH! n recommended flow of 0.25 mJ s_1 per rn2 of tank surface was adequate for some conditions and inadequate for others. An increase in flow above design recommendations did not necessarily improve hood performance. It appears that factors in addition to tank surface area need to be considered. Emissions to the workspace were only slightly correlated with local ventilation emissions, suggesting that fugitive emissions are related to factors other than hood flow’ rate. The authors suggested that these factors are probably more related to work practices. A more detailed analysis11 of the degreasing activities showed that workplace emissions were positively related to the number of times baskets or parts entered or exited the degreaser and the number of times parts or baskets were removed w’ith visible liquid solvent (carryout). Neither of these activities was correlated writh local ventilation emissions.
Lyiegbuniwe11 also looked at the effects of cross-drafts on hood efficiency. Cross-draft magnitude and direction near the vapor degreaser was measured at eight sites. The author was able to show that as cross-draft velocity parallel to the hood face increased, measured capture efficiency decreased (Fig. 10.9).
Cross-draft velocity was normalized by dividing the measured cross-draft. velocity by the capture velocity calculated at the tank centerline. Capture velocity at the tank centerline w’as calculated using Silverman’s12 centerline velocity I q. (10.1)) for unflanged slot hoods. There was considerable scatter in the data, show ing that cross-draft velocity alone is not responsible for low capture efficiency.
Where Q = hood flow rate (mJ s 1 or cfm); Vc = capture velocity (m s"1 or fpm); L =- tank length (m or ft); and X = 1/2 tank width if two slots are used (m or ft).
Changing Flow Rates
The velocities outside the exhaust opening diminish proportionally to diminishing flow rate. However, the efficiency of the exhaust could diminish more rapidly
1.2 |
0.8 |
|
*-> 0.6 <u 3 O. 5 0.4 |
0.2 0 |
0.0 1.0 2.0 3.0 Cross-draft velocity/capture velocity |
FIGURE 10.9 Hood capture efficiency versus normalized cross-draft velocity.1
Due to the changed relationship between exhaust air velocity and contaminant velocity. Increased flow rate usually increases efficiency, A large increase could result in drafts that are uncomfortable for the workers. Air velocities that are too high could disturb the process and result in increased material losses.
The supply airflow rate should approximately equal the exhaust flow rate. A minor difference between supply and exhaust flow rates should not disturb the exhaust, since exhaust systems usually are operated with higher pressure differences than supply systems. If the exhaust flow rate is higher than the supply, it could result in lower efficiency due to lower exhaust flow rates and cross-drafts (see Disturbances). If the exhaust flow rate is lower than the supply flow rate, there may be fewer problems with exhaust efficiency, but this could result in a supply airflow field different from the designed one and thus result in different kinds of disturbances.
Certain operations require that the workspace be at a lower pressure than surrounding workspaces, e. g., radioisotope laboratories. In these cases, the exhaust flow rate should exceed the supply flow rate, but this excess should be within 10%. The additional resistance resulting from this imbalance should be considered in the design of the exhaust system, specifically in the selection of exhaust fans.
Flow rate changes are sometimes used in the design of local exhaust systems. An example of this is the use of dampers, blastgates, or valves that are interlocked with the machinery of interest. When the tool is on, the damper is opened and air is exhausted from that hood. When the tool is off, the damper is closed and more exhaust air is available to other parts of the system. These kinds of systems diminish the cost for running the system but increase the cost for installation as well as periodic preventive maintenance. The supply system
must be run parallel to the exhaust system, otherwise there will be little to gain by the changing flow rates and the exhaust could be less efficient than desired.
All exterior hoods should be evaluated regularly. The evaluation procedures can be divided into detailed and simple procedures. Detailed procedures need special instruments and competence, whereas simple procedures may be performed daily. Since simple procedures do not directly measure the performance of the exhaust, it is usually necessary to calibrate them using detailed procedures.
Detailed Evaluation Detailed evaluation is performed by measuring the capture efficiency, either by using the actual contaminant or by using a tracer gas. (In principle, it is possible to use particles as tracers, but gases are usually used as tracers.) The most reliable evaluation is to use the process-generated contaminant, since there are always problems with a tracer, due to the difficulties of feeding the tracer to the source in the same way and in a similar amount as the generated contaminant.13
Capture efficiency is the fraction of contaminant generated captured directly by the hood. With the appropriate instruments it is not difficult to measure this efficiency. The total amount of generated contaminant can be found by either shutting off the local exhaust and measuring the concentration in the room or, preferably, by ensuring that all generated contaminant is captured during the concentration measurements in the duct. This can be ensured by putting the exhaust duct very close to or around the source or by making a temporary closed hood around the source. The captured amount is then calculated by comparing the measured concentration in the duct, when the exhaust is working in its normal position, to the measured concentration, when all contaminant is captured.
Capture efficiency can also be measured by first estimating workspace emission rates and local exhaust emissions. The local exhaust emission rate equals the duct concentration (mass/volume) multiplied by the duct flow rate (volume/time). The workspace emission rates can be calculated using appropriate mass balance models and measured ventilation rates and workspace concentrations. Capture efficiency is the ratio of duct emission rate to total emission rate (duct plus workspace).9’14
Capture efficiency could be measured in the same way with a tracer gas. The difficulties explained above could make this measurement less reliable even if it usually is easier to measure the concentration of a tracer gas than the concentration of a specified dust or gas mixture.
In theory it should be possible to calculate the capture efficiency without measurements. Some attempts have used computational fluid dynamics (CFD) models, but difficulty modeling air movement and source characteristics have shown that it will be a long time before it will be possible to calculate the capture efficiency in advance.15
Simple Evaluation A simple evaluation can be done by checking the airflow rate into the opening, presuming the source characteristics, the placing of the exhaust, and the other parameters (cross-draft, work routines, supply airflow rate, etc.) have not changed since the detailed evaluation was done. It is necessary to do the simple evaluation at the same time as the detailed evaluation. The flow rate into the exhaust opening can be measured in many different ways; measuring the pressure drop through the opening, using a flow measurement device in the duct, measuring air velocities in the opening plane, etc. Different flow measuring devices are described in, e. g., the ACGIH Industrial Ventilation Manual.10
Another simple way to evaluate the exhaust is to use smoke ro visualize the air streamlines. It is also possible to see how far the exhaust reaches by observing the smoke generated at different distances. This is best achieved by using the source (if it generates particles or warm gases) and a suitably placed lamp to illuminate the contaminants.
10.2.2.2 Basic Exhaust Openings
General
The basic exhaust opening (BEO) is the simple opening placed at the end of a duct or a tube, which acts as the connection to the suction device (fan). These types of exhausts are commonly used because they can be designed and operated in a way that does not interfere with the process. Most BEOs have quite simple forms—circular, square, rectangular, or nearly any shape—and they can also be tapered with different angles. A rectangular opening is called a slot when it has a large length-to-width ratio (>5). A slot can also consist of many smaller slots or holes in a line. BEOs can be provided with a plane surface covering parts of the opening (baffles). The plane surface could be provided with slots or holes to make a perforated surface. The intention of these surfaces is to improve air distribution over the opening, increase the velocities in the openings, or reduce exhaust flow rate (see Different Forms and Boundaries Relative to Other Types). A BEO could, instead of being one opening, consist of many small openings close to each other in the same plane. The opening plane for a BEO is most often straight, but it could be curved to fit the process or the tool.
Most BEOs are situated at the end of a tube, but there are also basic openings situated in walls. BEOs can be used for nearly all kinds of sources, but are usually used for point sources. Use for line or area sources usually demands flexible or movable exhausts, or a slot placed along the line source or along the sides of an area source, or a very large (circular or rectangular) opening placed close to the source. A high flow rate is needed to get efficient exhaust in many cases.
BEOs are used as separate exhausts, as parts of other exhaust systems, or in combined supply and exhaust systems. In these cases there are normally specific design procedures.
Basic exhaust openings are not recommended for use when the distance between source and hood is great, since it is easy for contaminants to spread outside the reach of the exhaust due to the sharp decrease in velocity with increasing distance. It is usually better to use partially closed systems.
Principle
The main principle in designing and operating BEOs is to generate an air velocity at or near the contaminant source or contaminant dispersion direction. This air velocity should be higher than the velocity of the contaminant and any disturbing cross-drafts and should be directed into the exhaust opening. The contaminant is thus transported from the source into the hood. The air velocity outside an opening decreases sharply as distance from the opening increases.
Therefore, the exhaust opening should be as close as possible to the source. The exhaust flow rate must be larger than the source’s volume generation rate.
Adjacent surfaces can act as barriers to airflow from uncontaminated areas and diminish the necessary flow rate. To prevent suction of air from the uncontaminated region behind the hood, the openings can be provided with flanges, which should be as large as possible. Another variation is to place the source close to surrounding surfaces. lb l7 BEOs can also be integrated info tools, e. g., as part of a frame for automatic welding by spot heating or as small holes in the grinding surface of a portable grinding machine (rotating or oscillating).
BEOs consisting of open tube or duct ends can be connected directly to a source. In principle, these could be called closed systems, since their main function is to exhaust contaminants directly from the source, which is enclosed in the duct. However, they are usually regarded as basic exhausts since they function as such when not connected to the source.
BEOs are also used as exhaust openings in closed volumes or as parts of exhaust systems for different supply systems (Section 10.3). In these cases, it is not the velocity distribution in front of the exhaust opening that is of importance, it is the airflow rate. These applications are described elsewhere (Section 10,4 ).
Figure 10.10 show’s velocity contours outside plain circular openings and outside circular openings w’ith flanges. Figure 10.11 illustrates what happens to the velocity contour outside a plain circular opening when it is placed on a surface. Velocity contours and the streamlines of freestanding circular openings
|
|
|
|
|
|
|
|
|
|
FIGURE 10.10 Velocity field outside circular (left) and flanged circular (right) opening. Velocities are given on a plane of the diameter and in percentages of opening velocity.1016
JHli FIGURE 10.11 Influence of adjacent surface on velocity profile for circular opening, i9 |
And a circular opening close to a corner are compared in Fig. 10.12. Figure 10.13 shows centerline velocities for circular, flanged openings according to different equations along with some experimental measurements.
Applicability of Sources
BEOs are most often used for point sources or small line or surface sources. See Chapter 7 for descriptions of sources. BEOs are sometimes used for lines or surfaces when the source is moving along the line or on the surface. This naturally demands the exhaust to move with (or be moved with) the source movements (e. g., during painting or seam welding). They have also been used for side suction from baths and tanks22 and these exhausts are usually called rim exhausts see Rim Exhausts. However, for these sources pushpull systems (Section 10.4.3) are often more efficient. Side hoods can also be used, e. g., when molten metal is poured; however, in these cases an enclosed exhaust is more efficient.
BEOs can be used for both warm and cold sources depending on available space. When used for warm sources (e. g., welding fumes or fumes from pouring molten metal) it is much better (usually necessary) to place the exhaust above the source, since the buoyancy force is strong enough to counteract an exhaust flow from below or on the side. See Chapter 7.5 for descriptions of airflow from thermal sources. Sometimes a source could make the contaminants move first upward and when reaching the rim move downward (e. g., during degreasing operations). Usually processes release contaminants in one main direction and the exhaust must be placed in a way that does not try to counteract these natural forces. Whenever possible, the
Freely suspended and when it is close to an angled wall.20 |
FIGURE 10.13 Velocity changes on centerline outside a flanged circular opening. Vertical axis is nondimensional velocity (real velocity divided by opening velocity) and horizontal axis is nondimensional distance (real distance divided by opening diameter). The lines are drawn according to different formulas and the circles are measured values.21 |
Air velocity of the exhaust should have the same direction as that of the contaminant generation (e. g., grinding).
In addition to the advantages of having the exhaust air move in the direction of the generation and the necessity of having a high enough velocity, the exhaust flow rate must be larger than the airflow rate induced by the contaminant generation. This induced flow rate is often many times larger than the flow’ rate of the generated contaminant (Chapter 7).
For exhausts connected directly to the source, the main problem is to get a tight connection between the exhaust and the source. A common example is exhaust of contaminants from car exhaust pipe when the car is inside a building. In this ease, the flow rate must be high enough to transport all generated car exhausts and the connection of the exhaust opening to the tailpipe must be tight enough to prevent car exhaust gases from entering the room air. Other examples are the exhaust around the handle used when pumping gasoline, the exhaust from a lathe machine with a readymade exhaust connection,10 and the exhaust from an electric arc furnace.
As mentioned above, a basic exhaust can be of nearly any shape with the most common being round, rectangular, or slot openings, with or without flanges. Openings in walls could be said to have the largest flanges. There also are exhausts consisting of multiple holes or perforated plates in a wall or ceiling or floor or table. The latter ones (holes in a floor or table) often have special designs and are treated separately in a later section. The exhaust opening can be tapered to have both a large opening area and a smooth velocity increase inside the opening to the connecting duct, resulting in lower pressure loss. As mentioned above, exterior hoods can be connected directly to a process or to equipment. These have some similarity to industrial vacuum cleaners.
A separate description also exists for low volume high velocity (LVHV; systems. Some of these are similar to BEOs, but they are used in a different way from basic openings. One difference is that BEOs have lower air velocities in the openings than LVHV systems, usually less then 15-20 m s"1 and more than 75 rn s-1, respectively. Another is that basic openings have opening areas larger than 0.01 m2, while LVHV openings have areas less than 0.001 rn2. Intermediate systems exist, such as the small circular exhaust opening around a welding rod or the exhaust in the surface of a portable grinding machine.
Specific Issues
The specific problems for BEOs are mostly related to the specific processes at which they are used. One problem that does not, directly, depend on the process is the use of the exhaust system. If an exhaust is not used, naturally it cannot remove contaminants. Although BEOs are less efficient than total or partial booths, their use is still justified. When the location of the BEO is not perfect, higher flow rates are usually required for contaminant control. However, it is better to use this type of hood than to have no local exhaust.
Another issue is the velocity distribution in the opening. Some equations describing the flow field outside a BEO presume that all points in the opening have the same velocity, equal to the flow rate divided by the opening area (Design Equations). This is usually a reasonable assumption for design purposes, especially if the opening is provided with baffles to get a smooth velocity distribution. In reality, the velocity distribution in an opening could be quite uneven, which can affect both the velocity distribution outside and the efficiency of the exhaust.
Some other problems include clogging of the opening with exhausted material and an opening shape that is not chosen to fir the process which results in use of higher flow rate than necessary, thus increasing the cost ot the process. Other problems are described in Design Considerations.
Design Equations and/or Parameters
The first step in designing an exhaust hood is to select the geometry of the hood. As described above, the hood should enclose the process as much as possible. Where enclosures are not possible the hood should be located as close as possible to the source. The next step is to select an appropriate hood flow rate. The most common methods are
A: Calculating contaminant and exhaust velocities at all points in the flow field.
B: Use of capture velocity and centerline velocity.
C: Use of “rules of thumb.”
For processes where gases or vapors are generated it is often possible to calculate both the velocity and flow rate of the contaminant, [’his could lie done with different degrees of accuracy by using chemical and physical principles. These are described generally in Perry et al.22 and Reid et al.2. Computer programs also exist that calculate source strengths for different kinds of processes and process equipment. By combining these values with the velocity distribution outside an exhaust it is possible to choose the appropriate exhaust airilow rate.
For processes that generate particles (dust, mists, fumes, etc.; there ;s no general procedure to calculate the contaminant’s velocity and flow rate. Data and figures exist for some processes describing the contaminant generation process; however, for dust-generating processes Method B or C is often used.
Method A: Calculating Contaminant and Exhaust Velocities at All Points in the Flow Field Local exhaust hoods are used to remove contaminants at the point of generation before they escape into the workplace air. The efficiency of any local exhaust system is greatly affected by the flow field generated by the exhaust opening. Therefore, accurate modeling of this flow field is essential for reliable predictions. However, solving the airflow field is a formidable task and often must be done numerically.
Potential Flow Airflow fields generated by unobstructed exhaust openings can be satisfactorily described with potential flow models. In these models the flow is idealized by assuming it is inviscid and irrotational, which greatly simplifies the mathematical treatment. This is possible because viscous effects occur appreciably only along solid surfaces and in the region close to the opening. These viscous boundary layers have little influence on the flow field of unobstructed exhaust openings and can therefore be ignored. Moreover, air velocities near exhaust openings in industrial applications are usually so low that the flow can be assumed incompressible.
For an irrotational, incompressible, and frictionless fluid flow there exists a scalar velocity potential 4> Such that the velocity vector V Is
V = V(j>, (10.2)
Or for a steady incompressible flow the continuity equation is
V-v=0. (10.3)
Substituting the conditions imposed by continuity gives Laplace’s equation:
V ■ V= V — V</> = V24> = 0. (10.4)
This equation, taken in conjunction with appropriate boundary conditions, can be solved analytically or numerically to give the ideal flow field.
The Stream Function Stream functions are defined for two-dimensional flow and for three-dimensional axial symmetric flow. The stream function can be used to plot the streamlines of the flow and find the velocity. For two-dimensional flow the velocity components can be calculated in Cartesian coordinates by
V = ^
L dx’
And for axisymmetric flow in spherical coordinates:
1 M
Risinfl3r |
Ur ‘ i ■ NRt*
0.6)
_ 1 91//
<8 Rsintidr’
Where U and V are the velocity vectors in the X and Y directions, respectively, If/ Is the stream function, R is the distance from the origin, and 0 is the polar angle measured from the. v-axis.
Two-Dimensional Flow
Line Sink The velocity induced by a line sink is purely radial and is given in cylindrical coordinates by
Wr=-f^, (10.7)
Kjrr
Where Q/L is the exhaust flow rate per unit length, R is the radius of a cylindrical control volume surrounding the sink, and K is a constant which has a value 2 for a line sink withdrawing air from the whole space and 1 for a flanged slot. The negative sign means that flow is toward the sink.
The stream and potential functions are
J^j-0 (10.8)
And
4> ~ — T — T In R. (10.9)
K ttL ‘
Flanged Slot Slots can be thought of as rectangular hoods with a very large aspect ratio (the ratio of slot length, L, to its width, W). Anastas and Hughes24 have shown that rectangular exhaust openings behave like slots when the aspect ratio is greater than 100. Usually an aspect ratio greater than 5 is used to define a slot.25 Slots are an important class of exhaust opening and they are typically used as rim exhaust on large tanks or welding exhaust hoods in high-velocity/low-volume local ventilation systems.
The analytical solution for an infinitely flanged slot can be obtained by assuming that the inlet is composed of elemental point sinks.26 The velocity field of an infinitely flanged slot can be obtained by assuming the velocity to be uniformly distributed across the opening. The contribution to the velocity potential at point (x, y) due to the elemental line sink of length DC, and located at (0, Q is given by
Wn U, -W/2 "7r |
nrd^, |
Where U0is the mean face velocity. The velocity components can be obtained by using Eq. (10.2): F"7/2 U0 x |
(10.13) |
Dx |
W/i N x— + (y-З) Y-C |
Hid |
V = ~ = -=e-1 —- In RdC = Dy Аyj Tt j-W/1 Tt x — + (y_ q |
D C |
(10.14) |
U(x, v) = ‘ TT |
Arctan |
L |
W |
Y + |
W |
+ "I X~ |
-^In |
V(x, y) = |
‘y-W/2^ |
Arctan |
W/2 |
In RdC, |
_w/2 TtLW |
/i |
R = i. v’ + i v — З) ) |
110.11 i |
The total potential at point (x, Y), due to all the elements in the flanged opening, is the sum of all such elements and is given by the integral |
|
|||
![]() |
|||
![]() |
|||
![]() |
|||
|
|
||
![]() |
|||
|
|||
![]() |
|||
|
|||
![]() |
|||
|
|||
![]() |
|||
![]() |
|||
![]() |
|||
|
The velocity contours calculated by this analytical model are shown in Fig. 10.15. The velocities are expressed as parts of the face velocity. In the same figure, velocities produced by the line sink model are also plotted, e. ir the exhaust opening the line sink model overestimates velocities but lurther from the opening the agreement is good: the experimental velocities are almost identical to the velocities given by the analytical model at distances X/W > 2 from the opening.
The analytical centerline velocities were calculated from
W 2x |
J7 Uo |
Ff- = — arctan 77 |
![]() |
|
|
|
Which is obtained from Eq. (10.15) by setting Y =0. It is interesting to note that when X/W is large, the centerline velocity in Eq. (10.17) can be approximated by the MacLaurin series expansion
U_ T’o |
X7t’ |
.10.18)
Which is the same as obtained by the simple line sink model.
Plain Slot The analytical solution for the slot in the two-dimensional case can be obtained by conformal transformation:27
Z = ew + w +1, (10.19)
Where W = (f> + Iip is a complex function of the complex variable z = X + iy:
X + Iy — e*e{* •Ґ <t> + ii}>+ 1 = E*( Cos I// + Isin Tj! r) + <^» + ji/r + 1. (10.20)
Separating the real and imaginary parts gives
X = c^cos l|/+<}>+^ (10.21)
4 |
2 y/W |
Analytical solution I ine sink
And
These equations cannot be used directly, and numerical methods are needed to compute the velocity components. The velocity components can be found by implicit differentiation and using an iterative technique.24
Figure 10.16 shows the calculated velocity contours as a fraction of the average velocity in the channel (average face velocity). In addition, velocities obtained from the line sink model are plotted. It can be seen that, compared to the line sink model, the calculated contours are displaced somewhat in the positive X direction, with the greatest relative difference near the exhaust opening and with decreasing relative difference as the dimensionless distance X / W increases.
Three-Dimensional Flow
Point Sink A point sink is defined as a point in space at which the fluid is continuously and uniformly drawn off. The radial velocity into the sink at a distance R from the sink is, in spherical coordinates,
Q (10.23)
Ki
Where Q is the flow rate through the sink and K is 2 for a point sink in a plane (infinitely flanged opening) and 4 for an unflanged opening. Since Q and R are positive, the negative sign indicates that the fluid flows toward the sink. Due to symmetry, the angular velocity U() — 0.
: 10.24) |
= JL1
K 7T T
; 10.25)
|
-~Я~~ cos 9. KiT |
Y/W |
(y-a)(z+ b) |
U(x, y, Z) = ~—r |
Arctan — |
C7(y~ «)2 + (z + B)- * (y — a)(z — B) "■ |
■ arctan — X |
X, Ky — a)2 + (z-b)2 + x2 j |
XJiy + A)2 + (z+ B)1 + x2 (y + A)(z — b) |
J(y + a)2 +7z — B)2 + x2 |
(y + a)(z + B) |
; — arctan — |
+ arctan — |
![]() |
(10.27
U 0 V{x, y, 2) = ~2irln |
And in the y direction:
Z+ b+ Jxr^+ (Y-a)2 + (2+ B)1 z-b+ Jx1 + (y + a)2 + (z~ b)1
2 + B + Jx^(y + Вџ~+(2+ B)2 Z — b + Jx2 + (y — a)2 + (z~ b)2^
(10.28)
Ј7, |
And in the 2 direction:
U/(x, Y, z) — — T^ln — Ht |
Y + a+ Jx1 + (y + A)2 + {z — B)2 Y-a+ Jx2 + (y—A)- + (z + B)2 j Y + A + Jx2 + (y + A)2 — K2 + B)2 y-A+ ,Jx^Ty-^^ b)2
(10.29)
Where A =L/2 is the half-length and B — W/2 is the half-width of the opening.
Along the hood centerline (y = 0 and 2 = 0), the velocity components in the Y and 2 directions disappear and the velocity can be calculated by
LW |
IxJax1 + L2 + W1 |
U 2 „ -rr = — arc tan Ur) TT |
![]() |
|
|
|
Flanged Circular Opening The velocity components for an infinitely flanged circular opening with a radius of R are29
-dadl |
(10.31) |
, , ^ __Јo |K f27r Lx
Uir‘X)—2”oo (x2 + r2 + I1 + 2r/cosa)3/2
Uo fK f2Tr V(r’X) = —2*LL : 2 |
Fe,±iЈos«j———— _dadl (1[)32?
0 J 0 (x~ + r" + L~ + Irlcosa)
The velocities can be calculated by integrating Eq. (10.31) and (10.32) numerically. To perform the calculations, the exhaust opening is divided into several area sinks and their effect on velocity at any point upstream of the exhaust opening is obtained by summing. For the centerline (r = 0), Eq. (10.32) yields along the X axis
/ X_ D / |
0.25 + |
|
|
|
|
|
|
![]() |
|
Where D (= 2R) is the diameter of the opening and x is the distance from the opening (Fig. 10.18).
(r, x)
FIGURE 10.18 Geometry for a flanged circular hood. |
Flanged Elliptical Opening The potential flow solution for an elliptical aperture in an infinite wall with a constant potential across the hood face is given by Lamb30 as
(]> — =f-S — fA————— DA————- /’10 14)
J 47tJ o [{a2 + )(b2 + A)A]1/2 ‘ ■
Where A is the positive root of
-J^ + — Ј— + Ј = I. (10.35)
A — + B~ + A ^
= = ^——- -IZE-^————- ’ (ia38) Where |
Equation (10.34) can be solved to give the velocity components in the X, y, And Z directions using the same coordinate system shown in Fig. 10.17:n
2ttE |
||
Qy(a2 |
+ A Y/2(b2 + |
A)3/2A3/2 |
2 ttE |
||
Qz(a2 |
+ A)3/V + |
A)1/2Aj/2 |
2ttE |
Y _ — Dx = Djfc _ Qx(a2 + A)i/2(b2 + A) ’/2A1/2 , j q ^ j |
And X is perpendicular to the hood face, Y is parallel to the hood length, and Z Is parallel to the hood width. The hood airflow is Q; a is the length of the major axis and B Is the length of the minor axis of the modeled ellipse. Laboratory measurements in the X, y plane (z = 0) showed Eq. (10.37) to be a good predictor of Y velocity. Agreement between measured velocity and velocity predicted by Eq. (10.36) was not as good and the authors provide an empirical correction for Eq. (10.36). The agreement between predicted and measured X velocity changed as a function of location in front of the hood, with better agreement further from the hood face. Assuming constant potential on the hood face results in predicted velocities equal to one-half the average velocity in the center of the hood face and infinite velocities at the edges. Experimental measurements indicate that potential flow underpredicts velocity at the edges of the hood and overpredicts near the center. An empirical factor was determined to force the slope of the predicted versus measured X velocity to be equal to 1. The factor is a function of A, which in turn is a function of X, Y, and Z locations. Equation (10.36) is replaced with
V = Dx = 24 = Q*(fl2 + A)?/2(62 + A)VV/2 lnA-11.8 rm 4(V Tt dx 2 tTE ‘ -10.7 ‘
Superposition of Flows Potential flow solutions are also useful to illustrate the effect of cross-drafts on the efficiency of local exhaust hoods. In this way, an idealized uniform velocity field is superpositioned on the flow field of the exhaust opening. This is possible because Laplace’s equation is a linear homogeneous differential equation. If a flow field is known to be the sum of two separate flow fields, one can combine the harmonic functions for each to describe the combined flow field. Therefore, if <pt and <i>2 are each solutions to Laplace’s equation, A(f>i + where A and B are constants, is also a solution. For a two-dimensional or axisymmetric three-dimensional flow, the flow field can also be expressed in terms of the stream function.
Flow Past a Point Sink A simple potential flow model for an unflanged or flanged exhaust hood in a uniform airflow can be obtained by combining the velocity fields of a point sink with a uniform flow. The resulting flow is an axially symmetric flow, where the resulting velocity components are obtained by adding the velocities of a point sink and a uniform flow. The stream function for this axisymmetric flow is, in spherical coordinates,
— Hr(s sin0)“ — t^-cosG, (10.41)
L k TV
Where s is the distance from the origin and Us is the velocity of the uniform flow. The model does not take into consideration the location of the unflanged exhaust duct relative to the streaming flow. However, for the flanged opening the only physically meaningful orientation is with the hood axis perpendicular to the flow.
The addition of a uniform flow with the sink flow creates a dividing streamline (Fig. 10.19), so that the contaminants released inside the dividing streamline would be captured while the contaminants released outside would escape. The
|
|
||
|
|||
|
|||
|
|||
FIGURE 10.19 Combination of a point sink flow field with a uniform flow field.
Stream function has value J’P = Q/(kv) at the dividing streamline and so its location can be found by expressing Eq. (10.41) in polar coordinates:
■: 10.42) |
X = — -2-
Kir 2 Kit
2q |
X2 , Xq |
( 1 K.7T U$X |
2 2 kirUs |
1 /2—j |
FnrUs |
2 kTrUs( |
2 q K ) |
1 « J |
1/2 |
This can be solved for R(x), giving32 R(x) = |
(10.43) |
This solution gives unequivocally the effective control range of both unflanged and flanged openings when the exhaust flow rate and velocity of the idealized cross-draft are known. The distance from the hood opening to the dividing streamline for a hood in uniform flow perpendicular to its axis is thus
1/2 |
K0) |
{ 2q I KvUs |
![]() |
|
|
|
This type of dependence of capture efficiency on the exhaust flow rate and cross-draft velocity has also been seen by Fletcher and Johnson8 who determined the capture efficiency of a flanged square exhaust hood in a cross flow.
Semi-Theoretical and Empirical Velocity Fields Since the use of formulas to calculate the velocities outside an arbitrary opening could be very tedious, only some examples of these formulae are given. These calculations are best done on computers and there are some dedicated programs to calculate and visualize the flow fields outside exhaust openings. There could sometimes be problems when calculating the velocity field outside an opening close to
Jz2Ta |
2 1 X+ + YZ + |
Jx1 |
+ y — |
V = |
1 |
1 + |
|
|
|
|
![]() |
|
|
|
|
|
|
|
|
|
|
|
![]() |
|
|
|
|
|
|
|
![]() |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
‘ y+ |
— arctan — |
Arctan — |
2 ■ Jz" + X2+ + Y X+ ■ Y+ |
Z ■ ^Jz~ + .V + Y’+ X — ■ y+ |
110.50) |
+ arctan — |
-arctan — |
Z ■ Jz2 + X+ + y: z ■ Jz2 + Xz + Y: The nondimensional velocity in the direction toward the opening is |
— Jvi |
+ Vv + IK |
= x — 0.5 № |
= X+0-5jT |
(10.52) Y^y+0-5jw y- = y-°-sJw And W and L are the extensions of the opening in the x and Y directions, respectively, and X, Y, and Z are the nondimensional distances along the different axes and arctan is the arctangent function. The nondimensional distances are defined as the real distances divided by the square root of the opening area. The opening area is equal to W times L. |
W |
X + |
+ Z" |
In |
2tt |
W 1 |
X — |
+ Z — |
![]() |
|
![]() |
|
|
|
|
|
![]() |
|
|
|
![]() |
|
![]() |
|
|
|
|
|
![]() |
|
|
‘x + f |
X-y |
Is V«, |
I 77 |
A rctan |
Arctan |
![]() |
|
|
|
|
|
|
Calculations are usually done using computational fluid dynamics (CFD).-’4’3—’ Empirical values for an unflanged circular opening are shown in Fig. 10.10. In Figs. 10.20 and 10.21, empirical values for different unflanged rectangular openings (1:1, 1:2, 1:3, 1:10) are shown.
Method B: Use of Centerline Velocity Models with Capture Velocity Capture velocity is defined as the velocity outside an exhaust necessary to capture the contaminant farthest away from the opening, when it has released its initial energy, and transport it into the opening. Selection of capture velocity depends on the source generation rate, speed, direction, and spread, as well as the effects of disturbances such as cross-drafts. Some problems with this design procedure are: generated contaminants usually do not have one single velocity, and the maximum release velocity is usually not known, nor are the temporal and spatial velocity distributions of the release. Some very approximate recommended capture velocities for different processes are given in Table 10.2.5′ It must be emphasized that these values have been the same since they were first published in the 1940s.10’16 Few results regarding capture velocity have been published. Most of these studies have shown that capture velocity is not a simple tool to use.8’58
Many equations and figures of velocities outside basic exhaust openings have been published. There are some summaries and comparisons of equations.34’39 These equations are empirical, theoretical, or semi-empirical descriptions, but they have all been thoroughly investigated and tested and they are reliable as long as the prerequisites coincide with the original descriptions.7 Most equations describe the velocity along the center axis of different opening shapes.
Equations (10.23) and (10.7) have been used as the basis for centerline velocity estimates. Usually these equations are modified empirically depending on the shape of the exhaust. The results are specific equations for circular openings, square openings, rectangular openings with different side relations, slots, etc. Further modifications are made when the hoods are operated with flanges. The modifications for flanges result in lower flow rates for the same capture velocity or higher capture velocities with the same flow rate.
Equations for centerline velocity outside circular, square, rectangular and slot exhausts without and with flanges are presented in Table 10.3. These equations have been chosen to have as large an application as possible. For very detailed calculations it is recommended that the original research references be consulted.37’39’40
For exhausts with flanges it is assumed that the flanges have, in principle, infinite width. This is only achieved approximately for an opening in a wall. If the flange width is larger than the square root of the opening area10’41 the centerline velocities have been found to be practically equal to the velocities for openings with an infinite flange. Velocities in regions other than on the centerline have not been examined in the same way, but velocities not far from the centerline should be influenced approximately in the same amount as on the centerline.
Method C: Rules of Thumb There are many “rules of thumb” for choosing BEOs. Many of these are connected to manufacturers, which
The sides. From top to bottom: square, rectangular 1:2, rectangular 1:3. The velocities are in percentages of the opening velocity and for a plane in each side’s centerline.18 |
1.6 |
Distance from opening [x/s) |
Distance from opening {x/s) |
FIGURE 10.2! |
Velocity contours outside a slot or a rectangular opening with a side ratio of 1:10. The |
Velocities are In percentage of the opening velocity. The top figure snows the velocities along the centerline of
(Reprinted by |
The short side. The bottom figure shows the velocities along the centerline of the long side.
Permission of John Wiiey & Sons, Inc.)
Through the years have gathered substantial experience regarding their own
Products. These rules are usually in the form of recommended airflow rates
For specific openings used for specific processes at specified distances and
Places. The ACGIH10 also provides design plates for a number of specific pro —
Cesses. Many of these design plates rely on the experience of users and also tall
TABLE 10.2 Capture Velocities for Various Industrial Processes In each category below a range of capture velocities is shown. The proper choice of values depends on several factors. Lower end of range, I. Room air currents minimal or favorable to capture, 2, Contaminants of low toxicity or of nuisance value, 3. Intermittent low production, 4. Large hood—large air mass in motion. Upper end of range, I. Disturbing room air currents, 2. Contaminants of high toxicity, 3. High production or heavy use, 4. Small hood—local control only.
|
Into the “rule of thumb” category. These data include recommended flow rates, expected pressure drops, and geometry specifications. Normally there are no specifications of the efficiency of the exhaust, even when it is possible to make such an evaluation.
The recommendation is to start with Method A, if possible. If this is nor possible Method C can be used, but it is recommended that the equations given in Method A, together with approximations of the contaminant’s properties, be used to check and/or verify the manufacturer’s data. Method B could be used if no other alternative exists. If the design is for a noncommercial or nonstandard exhaust, it is not possible to use Method C and it is strongly recommended to use Method A, even if only approximate values are available.
Pressure Drop
The calculation of the pressure drop for a chosen exhaust depends on the calculation method (Chapter 9). Pressure drop is usually calculated as the product of a hood entry loss factor, Fh, and the dynamic pressure in the connecting duct, Pj. The Pd is expressed as P ■ vl/2, where p is the air density and V is the air velocity in the duct. Some common hood entry loss factors are given in Table 10.4.10-16-36
The design equations for pressure drops can also be used for evaluation procedures (Section 10.5).
General
Rim exhausts are a specific application of slot hoods, which in turn are a type of exterior or capture hood. Rim exhausts are slot hoods placed along the rim or edge of an area source such as an open surface tank or vessel opening. Open surface tanks are widely used in industrial settings for cleaning, stripping,
Hood shape |
Centerline velocity |
Rйfйrences |
Circular, unflanged |
V 1 |
Burgess et al..4tt |
O 12.7 — Z~ + 1 |
Dittes et al. ‘ |
|
Circular, flanged |
V 1 |
Burgess et al.,4" |
Vo 8.^+1 |
Jansson ’4 |
|
V7 ~ centerline velocity, Vq |
= hood face velocity (Q /A),z = distance from hood face along |
|
Centerline / opening diameter |
||
Square, unflanged |
V __ 1 vn 8.6 • Z1 + 0.93 |
Burgess et al.4,1 |
Square, flanged |
V _ 1 vn 6.3 • Z1 + 0.25 |
Burgess er al.40 |
V — centerline velocity, Vu |
= hood face velocity (QA4>, Z = distance from hood face along |
|
Centerline /length of one |
Side |
|
Rectangular, unflanged |
V 1 |
Burgess et ai.,4" |
Braconnier3′ |
||
Rectangular, flanged |
V” H°-2s*(§n’’ [(u5*(0T |
Burgess et al.40 |
V = centerline velocity, Vn |
= hood face velocity (Q / (L W)), W = hood width, L — hood length. |
|
X — distance from hood face along centerline |
||
Slot, unflanged |
— = 0.27 ■ (—) ‘ V(, W |
Burgess et aL4!) |
Slot, flanged |
*< II O 0 53IX |
Burgess et al.4’1 |
V — centerline velocity, Vfl = |
= hood face velocity f Q / (L W)), W = hood width, L — |
= hood length, |
X ~ distance from hood face along centerline and in the longest plane |
Electroplating, and other coatings applications. Rim exhausts are amenable to open surface tanks because they do not generally interfere with the operations of the tank and usually will draw contaminated air away from the breathing zone of workers. This is in contrast to canopy hoods, which may also be used for heated open surface tanks, although these hoods will draw contaminated air through a worker’s breathing zone if the worker leans over the tank.
Principle
The rim exhaust is a source of suction that is placed along one or more sides of the area source. Air is drawn across the surface of the source and con taminated air is drawn into the hood. Specific examples of rim exhaust include open-surlace tank exhaust such as electroplating, cleaning, degreasing; table exhaust such as mortuary tables; and exhaust used during container filling such as barrel filling.
TABLE 10.4 Entry Loss Factors for Flanged or Unflanged Round, Square, and Rectangular Tapered Openings
|
The angle is the major angle on rectangular hoods.10 |
Applicability of Sources
Rim exhausts are suitable for area sources of contaminant. They are limited in the area over which they can draw with adequate velocity. In practice, the slot hood should be within 0.6 m of the far edge of the source. For an open surface tank this means that a slot hood on one long side is necessary for tanks up to 0.6 m in width; hoods on both long sides are necessary for tanks up to 1.2 m in width; and rim exhaust is not practical for tanks wider than 1.2 m. For those situations, push-pull ventilation or enclosure type hoods are recommended.25
Different Forms and Boundaries Relative to Other Types
Rim exhausts are slot hoods located on or around the edge of a source such as an open surface tank. Flanges may be added to decrease the airflow from behind the slot (uncontaminated air) and therefore increase the airflow from in front of the slot (contaminated air). The plenum downstream of the slot, if located above the slot, may act as a flange. Flanges may also be added to the sides of the source (tank) away from the slot hood. These flanges also act to increase the flow of contaminated air into the tank. Tank flanges, however, may interfere with process activities by limiting access to the tank.
Design Equations and/or Parameters
Rim exhausts, being one type of slot hood, use the same basic principles as given in the section on basic exhaust openings. The recommendation is to use the equations given in the Basic Exhaust Openings section for unflanged or flanged slot hoods or elliptical openings. The most common design method, however uses Method B, capture velocity. The design procedure involves selecting a capture velocity. The selection depends on the generation rate and toxicity of the contaminant as well as some consideration of disturbances near the local exhaust hood. For the case of open surface tanks, the generation rate and toxicity are usually combined to determine the class of contaminant. The class is then used to select an appropriate capture velocity. The ACGIH25 gives recommended capture velocities for a number of open-tank processes. Equation (10.55) is applicable:
Tank aspect ratio {WlL) |
Flanged or baffled |
Free-standing or unflanged |
0-0.09 |
1.0 |
1.5 |
0.1-0.24 |
1.25 |
1.75 |
0.25-0.49 |
1.5 |
2.0 |
0.49-0.99 |
1.75′ |
2.251 |
1.0-2.0 |
2.01 |
2.51 |
’A maximum flow rate of 1.27 m3 s 1 m 2 tank surface is considered Sufficient. |
Q = CVcLX, (10.55)
Where Q = hood flow rate (m3 s_1); C = coefficient, which depends on use of flanging and tank aspect ratio, see Table 10.5 (unitless); Vc = capture velocity (m s-1); L = tank length (m); and X = distance the hood must “reach” for contaminant capture, equal to tank width if one slot is used and 1/2 tank width if two slots are used (m).
The rim exhaust is placed on the longer side of the tank and the ratio of the width to length of the tank is the tank aspect ratio. Higher aspect ratios require higher hood flow rates per unit area due to the increased distance that the hood must reach. The highest hood flow rates per unit area would be expected for square hoods. Values of C are given in Table 10.5.25
The pressure loss associated with this type of exhaust opening is the sum of two pressure losses. The slot hood is usually thought of as a sharp-edged orifice and the duct entry (from the slot plenum) is a flanged opening. The recommended hood entry loss is given by Eq. (10.56):25
He = 1.78(VPstot) + 0.25(VPduct), (10.56)
Where Be = hood entry loss (pressure units); VPslot = velocity pressure in slot (pressure units); and VPduct = velocity pressure in duct (pressure units).
The ACGIH25 gives design criteria for several specific applications of rim exhaust including vapor degreasing, tables, mortuary tables, and barrel filling. Each of these design plates gives a recommended hood flow rate based on source surface area and has been determined based on some of the design considerations described above as well as practical experience with these types of hoods. As described previously, selecting hood flow rate based solely on source surface area can lead to decreased efficiency in the presence of crossdrafts, worker or process activity, or obstructions in the flow field.
Another design method uses capture efficiency. There are fewer models for capture efficiency available and none that have been validated over a wide range of conditions. Conroy and Ellenbecker2’5 developed a semi-empirical capture efficiency for flanged slot hoods and point and area sources of contaminant. The point source model uses potential flow theory to describe the flow field in front of a flanged elliptical opening and an empirical factor to describe the turbulent diffusion of contaminant around streamlines.
Potential flow theory is used to predict the velocity components (V*, V^„ V.) at any point (x, y, z) in front of an elliptical opening (see Flanged Elliptical Openings). A cross-draft can be added to the velocity components through simple vector addition. The velocity model has been validated in the laboratory. j2 Capture efficiency is calculated from the velocity model for the jcy-plane (z = 0) with an empirical term to account for turbulent dispersion of contaminant around streamlines. Equation (10.57) is the capture efficiency model:
(
Exp |
X-u
(O
Si 0.57; |
1 + exp |
(
10
K J
Where Ri = capture efficiency; X = distance, in the X direction, from the hood face to the point of interest; /x = empirically corrected distance, in the X direction, to the dividing streamline; and to = empirical parameter to account for the spread of contaminant around streamlines. The dividing streamline is the streamline that just enters the hood.
The spread parameter, &>, is calculated from
W = -0.15XC + 0.004 with Xc in meters. (10.58)
The empirically corrected dividing distance, /x, is calculated from
/x = 0.83XC. (10.59)
The theoretical distance to the dividing streamline, Xc, is a function of hood dimensions, hood face velocity, distance parallel to the hood face, and crossdraft velocity, and is calculated from the equations for flanged elliptical openings in Section 10.2.2.2.
Capture efficiency of area sources of contaminant release, such as open — surface tanks, can be calculated by numerically integrating the point predictions over the surface of the source.2 For open-surface tanks with a liquid level very close to the bottom of the slot hood, the source in front of the hood prevents air from being drawn from below the slot, causing more air to be drawn from in front of the slot. The effect is to increase the distance to the dividing streamline. The velocity field, modified by the presence of the contaminant source, is modeled by an image source of suction below the slot. This modification is necessary for plating, pickling, or cleaning
Tanks and for tables. With vapor degreasers, the liquid level is much lower
Than the bottom of the slot and does not interfere with airflow from below the slot. The assumption of an image source of suction is not necessary. This area source model was validated in the laboratory with very good results. An attempt was made to validate this model in the field. Capture efficiency of local ventilation on vapor degreasers was measured at 16 sites. Cross-draft velocities were simultaneously measured at 8 sites. Predicted and measured capture efficiencies did not compare well (Fig. 10.22).42 As discussed in the Disturbances section, there are many possible reasons for the poor agreement. The area capture efficiency model only considers hood flow rate and dimensions, source dimensions, and cross-draft velocities,
♦ |
♦ __________________________ |
………………………………… * . |
|
……………….. ♦…… ^………. |
♦ ♦ |
♦ ♦ |
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 |
1.2 |
0.2 0.4 0.6 0.8 1
Measured capture efficiency
FIGURE 10.22 Predicted versus measured capture efficiency of vapor degreasers under operating conditions.
Whereas actual capture efficiency also depends on work practices and activity at the tank.
Computational fluid dynamics methods may allow for more accurate predictions. These models account for turbulence and other parameters such as thermal effects. A description of these methods is included in Chapter 11.
10.2.2.4 Low-Volume/High-Velocity Exhaust Ventilation
General
It can be very difficult to control dust from some industrial operations, especially when particles are released at high velocities. Large particles can travel significant distances before slowing down enough for capture and control by conventional local exhaust ventilation. Fine particles may be caught in the rapidly moving boundary layer of air that develops with rapidly moving surfaces, such as grinding wheels. Fine dust particles can be conveyed within the boundary layer and be dispersed away from the point of generation. These characteristics make it difficult to control personal dust exposures.
This discussion will address needs, applications, performance characteristics, and design considerations for LVHV exhaust ventilation. The applications are primarily for dust control. LVHV systems can be effective for protecting workers from dust exposures and for recovering valuable process materials. The equipment, excepting the nozzles, involves technology that is the same as for large central vacuum cleaning systems.
LVHV nozzles require very careful design. They can be effective in capturing and removing dust from operations that are otherwise difficult to control, e. g., hand-held tools and some fixed-machine grinding and other operations. Installation costs for LVHV systems can be considerably higher than for conventional local exhaust systems. However, operating costs can be lower be
Cause of relatively small exhaust air volume flow rates and resulting lower costs of conditioning replacement air.
LVHV nozzles can create problems that may be sufficiently severe as to prevent their use, usually in the form of ergonomic encumbrances and excessive noise. These problems can be dealt with, to limited extents, and LVHV applications can be effective. It must also be understood that dust control by LVHV systems is ultimately limited. No ventilation control measure can ensure sufficient worker protection down to extraordinarily low acceptable dust levels. Worker protection must always be confirmed by industrial hygiene monitoring and evaluation, and administrative control measures such as respiratory protection may be necessary.
LVHV applications have not developed rapidly. Initial concepts were developed for industrial applications in the 1950s and 1960s. Refinements were made and commercial products have been available from the 1970s until the present time. Unfortunately, there has not been much development of new applications.
Certainly, some workplace operations involving highly toxic and/or valuable materials can be controlled more effectively by LVHV ventilation than by conventional local exhaust ventilation. These situations represent opportunities to improve worker protection, recover valuable materials, and to reduce replacement air requirements. Designers of local exhaust ventilation systems should be mindful of such opportunities and take advantage of LVHV control methods.
Principle
Low-volume/high-velocity (LVHV) local exhaust ventilation is an approach that can be effective in controlling high-velocity dust particles. LVHV ventilation involves careful design and placement of relatively small exhaust inlets, better described as exhaust nozzles. The exhaust nozzles are positioned very close, within
2.5 cm (1.0 inch) of the point(s) of contaminant generation. LVHV nozzles utilize high “face” velocities, typically in the range 50-100 m/s (10 000-20 000 ft min’1, fpm). High inlet velocities and close positioning enable LVHV nozzles to capture much of the dust that is generated, even when released at high velocities.
Applicability of Sources
The most common applications for LVHV ventilation are to control highly toxic and/or highly valuable dusts. These circumstances make it easier to justify the typically high installation costs of LVHV systems. Some of the earliest applications of LVHV ventilation were to control machining operations on radioactive materials and highly toxic metals and alloys, such as those involving beryllium.43“45 Hand-held high-speed rotating tools can be among the most difficult for achieving effective dust control. A number of specific LVHV applications have been developed for foundry and other dust-generating tools.25
If well designed and properly used, LVHV applications have the primary advantage of effective dust control. LVHV systems also require much less replacement airflow than conventional ventilation systems. This can result in significant savings in operating costs to condition the replacement air. These savings can help to offset the typically high installation costs for LVHV systems.
LVHV systems have disadvantages associated with installation costs, e. g., custom-made nozzles, expensive hose and duct, air-cleaning equipment, expensive air movers, and large electric motors (with high operating costs). Noise levels generated by LVHV nozzles can be hazardous to hearing. And, as for any engineering control, LVHV ventilation may not be sufficiently effective to eliminate potentially hazardous exposures. For example, despite a relatively long history of use with beryllium machining operations, some LVHV systems may not have been sufficiently effective to prevent chronic beryllium disease.46 The practical limits of LVHV ventilation are not well known. Control of dusts to microgram-per-cubic — meter levels (such as for beryllium) might be achievable by LVHV ventilation for some operations, but possibly not for other operations.
High-Speed Dust Control
Figure 10.23 illustrates a progression of “Before LVHV” and “After LVHV” dust control applications.47 Figure 10.23a shows a simple plain rectangular, wedgeshaped LVHV exhaust nozzle removing dust from a stationary grinding wheel. Figure 10.23B shows a prototype LVHV nozzle positioned on a hand-held grinding wheel. Figure 10.23c illustrates the effectiveness of a commercially-made LVHV nozzle on a hand-held grinding wheel. LVHV nozzles have been used to control dust from machining on asbestos-containing and structural plastic materials.40
Welding Operations Efforts have been made to use the LVHV design approach for controlling welding fumes. Sometimes, this can be an effective method. Sometimes, however, there can be serious problems with the high-velocity exhaust stripping away shielding gases and causing poor quality welds. It is also difficult for exhaust nozzles to survive without damage in industrial welding environments, where even relatively slight damage can cause significant changes in the high-velocity airflow patterns and adversely affect welding. Most successful point — exhaust applications for welding establish capture velocities lower than for LVHV dust control, but still higher than for conventional exhaust hoods.
Other Applications Very small, very low-flow, and relatively high-velocity exhaust inlets, similar to LVHV nozzles, have been used successfully to control fumes from electric soldering irons.48«49 Some investigations have been made into small, point-control exhaust ventilation for aerosols generated by high-speed dental tools. However, such low-volume point-control ventilation systems have not seen widespread use.
Design Equations and Parameters
Nozzle Performance Characteristics Relatively little has been done to prescribe design guidelines for these applications. Published guidelines consist largely of drawings and limited data originating from LVHV equipment suppliers.25’40’45’44’47
Research has been conducted to evaluate the effectiveness of several LVHV nozzle configurations for capturing tracer aerosols for simulated industrial operations.50 Studies have been conducted to evaluate dust capture for LVHV nozzles on hand-held tools, and have demonstrated over 90% collection efficiencies for operations such as grinding and sanding.51-53
Research has also characterized the performance of LVHV nozzles having simple geometric shapes.54“56 The experimental nozzles included seven profile shape variations for circular nozzles having a diameter (D) of 2.54 cm (1.0 inch)
|
After LV/HV |
Before LV/HV |
(<7> Simple rectangular-wedge nozzle near a stationary surface grinder |
(h) Roughly fabricated nozzle for a hand-held cup-wheel grinder |
|
-) Commercially made nozzle for a hand-held cup-wheel grinder
FIGURE 10.23 Applications of LVHV local exhaust ventilation for dust control.47
And eight rectangular nozzles, all having 6.4.5 cm2 {1.0 inch2) openings that were tested for plain and flanged profiles. The circular nozzle profile shape variations included plain (with “thick” and “thin” inlet walls), flanged, flared (tapered at 45°), rounded, plain wedge (45°), and rounded wedge (45°). Plain circular nozzles were also tested at different diameters: D = 1.27, 1.91, 2.54, 3.18, and 3.81 cm (0.5; 0.75; 1.0; 1.25, and 1.5 inch). The rectangular nozzles varied in end shape
From square (width-to-length ratio, WLR = 1.0) to rectangular (WLR = 0.5 and 0.25), and narrow slot (WLR = 0.10). These experimental LVHV exhaust nozzles were tested at different volume flow rates (Q), establishing calculated-average nozzle face velocities ( V0 = Q /A; A = inlet area) of 50, 75, 100, and 125 m s“1 (10 000, 15 000, 20 000, and 25 000 fpm), and at some higher values of V0.
Centerline Static Pressures Figure 10.24 shows centerline static pressure profiles (SPCL) versus X (cm Hg versus cm; inch Hg versus inch) for several experimental LVHV nozzles. Figure 1.0.24a compares plain, flared, and rounded circular nozzles. The plain nozzle experienced a pronounced vena contracta; the flared and rounded nozzles did not. The effects of a vena contracta can be severe for LVHV nozzles. Figure 10.24и Shows a progression of SPCL profiles for changing nozzle face velocity, V0. The vena contracta became more pronounced as V0 increased, ultimately showing abrupt discontinuities, caused by weak shock waves, when the airflow achieved sonic velocity at the throat of the vena contracta. The achievement of sonic velocity established aerodynamic choking (“critical” flow) in the nozzle flow. Aerodynamic choking has significant effects on nozzle noise, as discussed subsequently.
Nozzle Static Pressure Loss Overall nozzle static pressure loss (SPN) was tested for all of the experimental LVHV nozzles.55’56 Experimental testing has confirmed what would be expected, that nozzle shape and size variation can cause great differences in overall static pressure loss, especially at high airflow velocities. Figure 10.25a compares SPN versus V0 (cm Hg versus m s_l; inch Hg versus fpm) characteristics for five circular nozzles. The plain wedge had the steepest rising curve, followed by the plain circular nozzle. Roth of
|
15.2 |
5 J2 1u -Ы Ј |
P-, |
(-4) |
(-2) X, cm (inches) |
(2) |
(a) Plain, flared, and rounded circular nozzles (D = 2.54 cm, V0 = 100 m/s)
(b) Plain circular nozzle (D = 2.54 cm)
FIGURE 10.24 Centerline static pressure (SPCL) profiles for experimental LVHV exhaust nozzles.57 The centerline pressure is measured at different distances, X, from the opening plane of the exhaust nozzle, X is positive outside the opening.
25.4 |
|
(10) . |
’ P-wedge —i II . |
20.3 |
■ Plain —11 11 |
Ofi / a ‘i |
|
Flared — / / L! l |
|
Vr Ij! 5.2 |
■ R-wedge —> Iff |
■s (6) |
Rounded Y X /// |
Ј |
|
10.2 ■ |
|
Ј (4) . |
|
‘J’. |
|
5.1 ‘ |
|
(2) . 0 |
50 100 150 (10,000) (20,000) (30,000) Vo, m/s (fpm) |
O 2.8 5.7 8.5 (100) (200) (300) Q, mVmin (cfm) |
(a) Circular nozzle profile shapes (D = 2.54 cm)
Hi FIGURE 10.25 Nozzle static pressure loss
Nozzles.55
(b) SPN vs. Q for plain circular nozzles of different diameters
Characteristics for circular experimental LVHV exhaust
These nozzles experienced substantial, irreversible energy losses caused by turbulence in the vena contracta. The flared, rounded, and rounded wedge profiles did not experience pronounced vena contracta effects, and experienced consequently lower values of SPN. Figure 10.25 b shows SPN versus Q (cm Hg versus m3 s“1; inch Hg versus cfm) for five plain circular nozzles having different diameters; the curves show large differences. Nozzle size has a profound effect on nozzle static pressure loss (SPN). This may prevent the use of very small LVHV nozzles that can experience extremely large static pressure losses.
Noise Characteristics LVHV exhaust nozzles can generate very high noise levels.35,56 Nozzle noise can present significant hazards to hearing. Options are limited, but actions can be taken to help reduce nozzle noise. Figure 10.26a compares nozzle noise profiles, dBA versus V0 (m s-1, fpm) for plain circular, square, and rectangular (WLR = 0.25) nozzles. The curves are remarkably similar, each showing dramatic increases in noise levels as V0 approached maximum flow, and each showing significant dBA noise reduction when driven closer to maximum (choked) flow.
Figure 10.26b shows noise spectra (dB versus Hz) for a plain circular nozzle at different face velocities, V0, As V0 increased, so did the noise levels, especially at frequencies greater than 1000 Hz. This was true at the flow rate causing the maximum noise level, i. e., V0 = 137 m s_1 (27 000 fpm). However, further relatively slight increases in flow rate (increasing V0) resulted in overall noise levels dropping dramatically (10-15 dB), especially in the highest frequencies (greater than 8000 Hz). These findings confirm that nozzle noise results largely from high-frequency airflow turbulence. The data show that driving an LVHV exhaust nozzle to its maximum flow, i. e., toward aerodynamic choking, can also
Noise level. dBA |
60 |
100(20,000) 25-75 — |
‘(5,000-15,000) Niiiml ‘ I 11utwf— |
10 10 10J 10^ |
50 |
|
|
|
|
![]() |
|
|
|
![]() |
|
|
|
|
|
|
|
(a) Noise profiles for plain circular, square, and rectangular (WLR = 0.25) nozzles (V0 = 100 m/s)
(b) Noise spectra for a plain circular nozzle (D = 2.54 cm)
|
85 dBA 90 dBA |
95 dBA 100 dBA |
(c) Noise potentials for a plain circular nozzle (D = 2.54 cm, =100 m/s) |
(D) 90 dbA noise potentials for plain, plain-wedge, and rounded-wedge circular nozzles (D = 2.54 cm, Vo = 100 m/s) |
FIGURE 10.26 Noise characteristics for experimental LVHV exhaust nozzles.51-56
Result in large reductions in nozzle noise levels. Less noise is propagated outside the nozzle because much of the noise is, in effect, “swallowed” by the nozzle.
Noise levels for plain circular nozzles propagated for greater distances as diameter increased for D = 1.27,1.91, and 2.54 cm (0.5, 0.75,1.0 inch), but were essentially unchanged for the larger sizes, D = 2.54, 3.18, and 3.81 cm (1.0,1.25,
1.5 inch). These findings suggest a limiting effect of nozzle size on nozzle noise. This would favor using smaller nozzles to reduce noise, but higher pressure losses will also occur. Increasing nozzle size beyond certain values (e. g., D = 2.5 cm for a plain circular nozzle) may not significantly increase nozzle noise levels.
Figure 10.26 also presents comparisons of nozzle noise potentials, two-dimensional plots of constant noise (dBA) levels in proximity to the experimental LVHV nozzles. Noise potentials for a plain circular nozzle.
Fig. 10.26c, demonstrated a progression (85, 90, 95, 100 dBA) of roughly elliptical contours, resembling constant velocity contours. The possible effect of LVHV nozzle noise projecting to great distances is illustrated in Fig. 10.26d (note the change in scale to meters (feet) from centimeters (inches)). The plain wedge circular experimental nozzle was extremely loud, establishing a 90 dBA noise potential encompassing an area about
3.8 m x 3.3 m (15 feet x 13 feet)—this area being roughly 50 times that for either the plain or the rounded wedge circular nozzles.
The extreme noise of the plain wedge nozzle resulted from a highly turbulent and obstructive vena contracta forming in the flow, downstream of the short side of the nozzle opening. This example highlights the possibility of entirely unacceptable noise levels from sharp-edged LVHV nozzles, and the significant benefits (noise reduction) that can come from rounding the inlet edges of LVHV nozzles.
Aerodynamic Choking Aerodynamic choking is an important factor to consider in the design of LVHV nozzles. Choking can occur whenever a nozzle is connected to a system capable of generating negative static pressure in the flow low enough (roughly 53% of atmospheric pressure; see Fig. 10.246) to cause “critical” flow through the nozzle. Choking is important because it limits nozzle flow rate and because it can reduce nozzle noise.
When choking occurs in conjunction with a turbulence-caused vena contracta, such as for plain and flanged nozzles, the overall nozzle static pressure losses (SPN) will be quite high. If choking were to occur in conjunction with a carefully designed converging-diverging nozzle wall, then it should be possible to reduce the obstructive turbulent region significantly, thereby reducing static pressure losses and noise levels. Such design could also have the advantage of achieving aerodynamic choking at more predictable flow rates. A converging — diverging type LVHV nozzle probably would require fabrication as a cast metal or as a molded plastic. Despite the potential benefits, nozzle design to facilitate aerodynamic choking has not been developed in LVHV applications.
System Design Considerations
The primary components of LVHV ventilation systems are the exhaust nozzles, flexible hose, fixed duct, air cleaner, and the exhauster and motor. Each of these is discussed below. Figures 10.23,10.27, and 10.28 illustrate LVHV nozzles. Figures 10.29 and 10.30 illustrate other system components and installations.
LVHV Nozzles Some of the earliest LVHV nozzles were developed for grinding and other high-speed, dust-generating, foundry-type operations. Handheld tool applications present special opportunities for LVHV applications. Figure 10.25 shows published nozzle designs for cup-shaped, disc, and cone-shaped grinders, a vibratory sander, and a pneumatic chisel.25 These illustrations are taken from several ACGIH “VS-Prints” and are based on nozzle designs that have been made commercially. The original British and United States patents for designs similar to some of these provide additional details and information,57-60
Figure 10.28 shows a complete ACGIH VS-Print for an LVHV nozzle applied to a surface-grinding machine. The figure illustrates an adjustable LVHV nozzle positioned near the point of grinding. Additional details are provided
|
Adapter plate to fоt grinder |
Hood adjustable for wheel Wear
{a) Cup-shaped grinder and nozzle design details
|
(b ) Disk grinder |
Sluts |
{c) Cone-shaped grinder T
|
(d) Vibratory sander (e) Pneumatic chisel
FIGURE 10.27 Examples of LVHV exhaust nozzles for dust control on hand-held tools.15
Ergonomic problems can cause significant challenges for LVHV applications on hand-held tools. The weights of flexible hoses can greatly alter the balance of tools fitted with LVHV nozzles. Positioning of the tool and nozzle for best dust capture can limit and encumber a tool operator. Noise controi options are usually limited to reducing sharp corners and controlling volume flow rate. Some findings suggest that molded plastic nozzles can dampen noise propagation, and thereby generate lower noise levels in comparison to fabricated metal nozzles. Molded nozzles could also incorporate converging-di — verging wall surfaces to facilitate aerodynamic choking.
Flexible Hose There are practical limitations on the size (inside diameter) and length of flexible hoses to connect from LVHV nozzles to fixed duct systems. LVHV hose must be smooth and flexible, strong enough not to collapse under low static pressures, and sufficiently durable to withstand industrial environments (e. g., dragging, flexing, impact, temperature variations, abrasive dusts). LVHV hoses are typically made of plastic or rubber materials, but should never be of the corrugated type. Conductive, static-proof hose, or bonding connections may be necessary to reduce accumulation of static charge.
LVHV hose inside diameters are usually recommended to be in the range 2.0-3.5 cm (1-1.5 inches).25 Hose lengths should be limited to about 2 m (810 feet), if possible. If greater lengths are necessary, then the hose inside diameter must be enlarged to reduce friction losses, but not so large as to fail to transport dust to the duct system. It is advisable to obtain accurate hose friction loss data from manufacturers.
Fixed Duct/Tubing The ducts for LVHV systems ordinarily consist of steel tubing designed to minimize turbulent pressure losses and convey airborne dust to an air cleaner. The type of steel-tubing duct that is used in cen tral vacuum cleaning systems is good for LVHV applications. It is designed with long-radius turns and smooth inside joints between the fittings and straight duct sections. The duct fittings have expanded ends to allow a slip-fit with the straight duct and a smooth joint on the inside surface of the duct. The slip-fit components are positioned and welded to create a strong and smooth airflow system.
Materials other than steel are also available, such as galvanized steel, aluminum, or stainless steel. Solid neoprene fittings are available for applications involving highly abrasive dusts.47 Sometimes, heavy cast iron pipe can be used for highly abrasive applications.25 Conventional sheet-metal duct cannot be used because it cannot handle the very low static pressures. LVHV and central vacuum cleaning system duct manufacturers should be consulted to provide accurate data for duct friction and turbulence pressure losses. Data available for conventional galvanized duct and fittings may not be sufficiently accurate for LVHV design applications.
Figure 10.29 illustrates duct sizing for a five-nozzle LVHV system, ft also illustrates connections to air-cleaning equipment and a multiple-stage centrifugal exhauster. It should be noted that the actual LVHV duct is much smoother than is suggested by the drawings in Figs. 10.28 and 10.29.
Air Cleaner/Dust Collector There are a wide variety of possible types of air-cleaning equipment for LVHV systems. The most common is a fabric dust collector, having a vertical cylindrical baghouse design. This type is common in large central vacuum cleaning systems. Other types of fabric collectors can also be used. For applications involving substantial amounts of large dust particles, it is usually a good idea to have a cyclone-type primary separator upstream of a fabric collector.
Air cleaning (dust collection) can be cost effective for LVHV systems handling valuable dusts. Care must be taken when handling potentially toxic dusts from air cleaners. Regular, routine reconditioning of fabric filters (e. g., by automatic shaking or pneumatic pulsing) is important. This can be accomplished on a set maintenance schedule or as a function of pressure drop across the fabric filter. It is not recommended to recirculate airflow back to the workplace because of the low air volume and potential hazards in the event of filter failures.
Exhauster and Motor Air movers for LVHV systems are not conventional fans. The low static pressures needed to operate LVHV systems can be generated by multiple-stage centrifugal (turbine-type) exhausters. These utilize high-precision rotating blades that can be damaged by dust. Consequently, it is always necessary to have an air cleaner in an LVHV dust control system to protect the fan. The low static pressures also require air volume flow rates to be corrected to standard conditions for exhauster selection.
The exhauster and motor must be mounted on a rigid frame. Installation should include a solid foundation, with vibration isolation of the exhauster and motor frame, and flexible connections between the inlet and outlet ducts and the exhauster. A commercial silencer should be considered to reduce noise levels propagating from the exhauster discharge stack. Electric motors for LVHV systems may need to be quite large in order to drive the exhausters at the necessary speeds and pressures. Motors can be connected by direct-drive couplings or indirectly by pulleys and belts. The operating costs of large motors can be significant, partially offsetting savings from low airflow rates.
Figure 10.30 illustrates a multiple-stage exhauster and smooth-flow duct (pneumatic tubing) components. It also includes pictures of air cleaner — exhauster-motor installations located outside of buildings and connected to LVHV (or central vacuum cleaning) systems inside the buildings.
General
Receptor hoods, also called canopy hoods, are designed to capture contaminants given off by heated processes. They take advantage of the thermal updraft caused by such processes; by placing the hood in the path of the updraft, they “receive” the exhaust and capture the contaminants.
Principle
Heated sources can cause strong updrafts that carry contaminants upward. Receptor hoods take advantage of this updraft, as shown in Fig. 10.31. The process shown to the left in Fig. 10.31 is at room temperature, while the process to the right in Fig. 10.31 is operating at elevated temperature. A canopy
|
FIGURE >0.31 Canopy hoods over a cold process (left) and a hot process (right).
Hood is shown over each process. The hood to the left is a simple exterior hood (see Section 10.2.2.2); as such, it must create a sufficient capture velocity at the surface of the process to capture the contaminants emitted. The hood to the right, by contrast, utilizes the updraft created by the process to aid in the contaminant capture; here the key design variable is not the capture velocity at the process surface, but the total volume of air set in motion by the updraft. As shown below, canopy hoods work well as receptor hoods on heated processes, but are very poor choices for room-temperature applications.40
Applicability of Sources
The key variable in determining the applicability of a receptor hood to a particular source is the temperature of the heated source, and the resulting updraft. The temperature must be high enough to cause an appreciable updraft, or the hood will be ineffective. An estimate must be made of the total amount of buoyant airflow set in motion by the heated source; the airflow through the hood must be greater than this buoyant airflow, in order to ensure complete contaminant capture. This principle is illustrated in Fig. 10.32, which shows the air spill that occurs when a hood’s exhaust airflow is less than the thermal updraft airflow.
Different Forms
Hemeon61 divides receptor hoods into low hoods, located within about 1 m of the heated source, and high hoods, located beyond this distance (Fig. 10.33).
FIGURE 10.32 Canopy hood with an airflow rate less than the thermal updraft airflow from a hot process. |
|
The design of low hoods is much simpler, since the adverse effects of turbulent mixing and cross-drafts are much less important than for high hoods. Low hoods are much more likely to capture a high percentage of the heated air and contaminants than high hoods, so they should be used whenever possible.
Specific Problems
The location of a canopy or receptor hood above a process, in order to take advantage of the thermal updrafts, can be problematic.40 If the top of the heated source is located below the worker’s breathing zone, as is frequently the case for open-surface tanks used in plating, it is very easy for the worker to place his or her head directly into the path of the rising contaminants resulting in very high contaminant exposure even if the hood is 100% efficient at capturing the source emissions. This is illustrated in Fig. 10.34. This represents a serious limitation on the use of receptor hoods; if workers must be positioned over heated sources, it is usually better to forgo the benefits of the thermal updraft and use an alternative design, such as a slot hood (Fig. 10.35).
Design Equations and/or Parameters
Hemeon61 is the first standard reference book that presents the equations for calculating thermal updrafts. These equations are repeated and expanded in other standard reference books, including Heinsohn,36 Goodfellow,16 and the ACGIH Industrial Ventilation Manual.25 These equations are derived from the more accurate formulas for heat transfer (Nusselt number) at natural convection (where density differences, due to temperature differences, provide the body force required to move the fluid) and both the detailed and the simplified formulas can be found in handbooks on thermodynamics (e. g., Perry22, and ASHRAE62).
High Receptor Hoods The important variable that distinguishes receptor hoods from other exterior hoods is Qus, the upward airflow set in motion by the heated source. Let us first consider the more general (and difficult) case of a high hood. Assume for simplicity that the source and the hood are circular in cross-section. The basic geometry used in this case is shown in Fig. 10.36.
According to Heinsohn:36
Qvs = O^z1-67*0’33, (10.60)
Where
Qt№ = buoyant airflow (m3s_1),
Z = virtual source height (m),
F = buoyant flux parameter (m4 s-3)
As shown in Fig. 10.36, the plume expands as it rises. The virtual source height is obtained by extending the actual source downward to the virtual point source:
Z — h + h’ , (10.61)
Where
B = distance between the actual source and the hood (m)
B’ = distance between the actual source and the virtual source (m)
FIGURE 10.34 Basic geometry for calculating necessary flow rate for high canopy hoods. |
FIGURE 10.35 Inappropriate use of a canopy hood, since worker exposure is not prevented by use of the hood. |
■H FIGURE 10.36 It is better to use a slot hood when a worker needs to lean over a bath. |
Several different equations for H’ are given in the references and are shown in Table 10.6.
Given this range of values in the literature, it is probably easiest for the user to go to the original source61 and assume that H’ = 2Ds.
The buoyant flux parameter, F, is often very difficult to determine. Theoretically,
F =-Ј&—, (10.62)
CP * oPo
Where
G = acceleration due to gravity (9.81 m s~2)
<f> = convective heat transfer from the source to air (W) Cp = specific heat of air (J kg-1 K-1)
T0 = absolute temperature of ambient air (293 K) P0 = density of ambient air (1.18 kg m-3)
The convective heat transfer is given by
® = hcAs(Ts-T0), (10.63)
Where
Hc = convection heat transfer coefficient of the heated surface (W m 2 K-1) As = heated source surface area (m2)
Ts = heated source surface temperature (K)
The difficulty in using these equations lies in the determination of Bc for any
Given heated surface; in general, the value of Ht. will not be known since it de-
TABLE 10.6 Equations for Distance between the Actual Source and the Virtual Source, H’ (m)
|
D, — diameter of the heated source (m). |
Pends on the specific geometry of the surface. Hemeon61 presents equations for calculating Hc for several simple geometries, including horizontal and vertical plates and cylinders; these equations are given in Table 10.7. {The equations given in Section 7.5 could also be used for calculation of plume diameters.)
The solution to the above equations will result in a value for Qvs, the airflow set in motion by the heated source. The actual airflow through the hood,
Qvli, must be larger than Qvs to ensure complete contaminant capture.
Heinsohn36 recommends that
QuH=.2qvs. (10.64)
The diameter of the receptor hood is also a critical design variable. The diameter of the heated plume, Dp, can be determined geometrically if it assumed that the included angle of expansion is 18°. Alternatively, ACGIH25 and Goodfellow16 give the following equation for the plume diameter:
Dp = 0.5zOM. (10.65)
TABLE 10.7 Heat Loss Coefficients (hc) by Natural Convection43
Shape or disposition of heat Natural convection heat loss
Surface coefficient (W m’1 K_l)
Vertical plates over 0.5 m high 2.0 (ii,.)0 25
Vertical plates less than 0.5 m high 1.4 {ATr/ L„)° 25
[Lv — height in meters)
Horizontal plates facing upward 2.5
Horizontal plates facing down — 13 (At,.)0-25
Ward
Single horizontal cylinders 1.1 (Afr/Z^)0-2-5
{Lt! = diameter in meters)
Vertical cylinders over 0.5 m high 1.05 (Ai^/ Lj)0-25
(Ld — diameter in meters)
For high hoods, the plume may be deflected by cross-drafts, and the canopy must be larger than Dp to account for this effect. Hemeon61 gives no specific procedure for calculating the hood diameter, stating only that the hood diameter should be “materially larger” than the plume diameter. Heinsohn-56 recommends that the hood diameter be equal to one-half the effective height (z). ACGIH25 recommends that
Df= dp + 0.8 k, (10.66)
Where Df = diameter of hood face (m).
Although the above approach for high receptor hoods is rigorous theoretically, it is difficult to use in practice. ACGIH2j offers a simpler approach.
First, the virtual source height is calculated using Eq. (10.61) where
B’ — 2d’/138. Rather than calculating QvH, the velocity of the hot air column at the hood face is calculated using
Pf=0 ;85A^gp7^t (10>67)
Z
Where
C“1
Vf = upward plume velocity at the face of the hood (m SAs = area of the heated source (m2) x = height of the plume (m)
Ts — T0 = difference between heat source surface temperature and ambient air temperature (K)
The diameter of the hood is then calculated using Eqs. (10.65) and (10.66). An inward velocity, V„ must be selected for the air flowing into the hood through the parts of the hood without the plume; Hemeon63 recommends a velocity of 1 m s_1, while Heinsohn36 suggests 1.5 m s-1.
Finally, the required hood airflow rate is calculated as the sum of the plume airflow and the inward airflow through the rest of the hood:
QvH = VfAp + Vr{Af-Ap) , (10.68)
Where
QvH = total airflow through the hood (m3 s_1)
Ap = area of the hot air plume at the hood face (m2)
Af = total area of the hood face (m2)
Vf = 1 or 1.5 (m s_1)
This approach is much easier to use than the first, but may give less accurate results since it assumes an average or typical value for the heat transfer coefficient for all heated source geometries.
The same basic procedures as described above can be used if the source and hood are rectangular rather than circular. The two dimensions of the plume and hood must be calculated independently by substituting them for Ds In the equations developed above.
Low Receptor Hoods Low receptor hoods are much easier to design, since entrainment of air into the plume and the effects of turbulent crossdrafts are not significant problems. In this case, the diameter of the plume at the hood face, dp, is assumed to equal the diameter of the source, Ds. Accord
Ing to ACGIH25, the diameter (or side length, for a rectangular hood) need only be 0.5 m larger than the source. ACGIH25 gives the following equation for the airflow into a low circular receptor hood:
QvH = 0.045djSi(Tf — T0)°’425 (10.69)
Where
Dp = diameter of plume at hood entrance (m)
Ts — T0 = difference between heat source surface temperature and ambient air temperature (K)
QvH = flow rate into hood (m3 s-‘1)
If the source and hood are rectangular, the following equation is used:
QvH=OMbi33L(Ts-To)0A1, (10.70)
Where B = the width of the rectangular hood (m), L = the length of the rectangular hood (m), and QuH = flow rate into hood (m3 s~]).
Evaluation
Capture efficiency measurements may be used to evaluate the function of a canopy hood (see Section 10.5). Capture velocity is not a feasible evaluation tool, since a canopy hood does not generate an air velocity close to the source. It is also possible to use exposure measurements for workers outside the plume area. Since most hot processes generate visible contaminants, visual inspection of the flow, especially around hood edges, might provide a qualitative evaluation. Many contaminants could however be invisible when diluted and smoke generators (Section 10.5) may be necessary to find leakages (temporary or permanent) around the hood edges.
10.2.2.6 Downdraft Ventilation Tables
General
For some operations it is advantageous to have the exhaust downward instead of through one of the basic exhaust openings described previously. This is accomplished by suction through the working table, a downdraft table. These exhaust openings are quite like basic exhaust openings with flanges, directed upward. To make the surface function both as an exhaust and as a working table, the opening of the exhaust is covered by a perforated table. To this table could be added vertical and horizontal walls and possibly a ceiling, which makes such an opening more like a partial enclosure than a basic exhaust opening.
Principle
Slots or holes are placed into a horizontal work surface to exhaust air downward in such a way that the surface can also be used as a table. The amount of open area for airflow varies depending on use.
Applicability of Sources
Small versions of downdraft tables (less than approximately 0.5 m2) are used when small-sized chemical work is to be done on tables instead of in laboratory fume hoods (see Fig. 10.37). This includes work with low-momentum sources (no initial velocity and near room temperature) such as laboratory animal experiments.
FIGURE 10.37 Downdraft hood for small-scale laboratory work. |
Downdraft tables are also used for human dissection tables although rim exhausts (see Rim Exhausts) are usually used for this purpose. For manual shoe repair and during application of artificial fingernails, downdraft tables are the only type of exhaust possible.
Larger versions are used for sanding, grinding, polishing, and welding. Since welding generates large amounts of heat together with the contaminants, a downdraft table, although handy, is not suited for this use. There could be problems with grinding, since the contaminants are generated with a high velocity in different directions and a downdraft table is only suited for capturing contaminants with low velocity and a temperature equal to or lower than room temperature. A partial solution to this is to surround the table on three sides with walls (0.2 to 0.5 m high) to deflect the generated contaminants and to diminish the amount of air pulled in from the sides (see Fig. 10.38).
Very large downdraft tables have been used for sources such as electrocutting and welding of large steel sheets.
Different Forms and Boundaries Relative to Other Types
For small-scale laboratory work, the exhaust surface is often made as a separate section added to the side of a table or put into a large hole in a table. These tables usually have a sheet metal surface that is resistant to the chemicals used and is easily cleaned. Many circular holes are cut into the metal surface to allow for airflow. This perforation makes the pressure difference over the table quite high and at the same time gives an even distribution of the airflow over the entire surface. These types of exhaust surfaces could be formed to suit different working conditions, e. g., the surface could be made to fit into a sink or to be placed below and around a balance. Using side walls that are not too high, on three or four sides, transforms the table to a partial enclosure, which increases
Tiors to distribute the flow over the entire surface. |
The capture of the contaminants. When a downdraft table is surrounded by three walls and a ceiling its function is more like a booth (see Section 10.2.3.2).
For sanding, grinding, and polishing, both small and large tables exist that use quite large flow rates. These often have built-in fans and filters, which could either return the air to the workroom or blow it to the outside. These tables could be placed into larger working tables but more commonly separate fixed or movable units. They usually do not have a sheet metal surface; instead, the exhaust opening is covered with metal (steel) bars for heavy loads. They could also have a collecting tray beneath the surface and could be surrounded on three sides by vertical shielding walls as mentioned above.
Downdraft tables are available for welding, but the efficiency could be quite low due to low suction velocities and large work pieces.
Similar types of tables are used for electro-cutting of large steel sheets, where it is common to divide the exhaust into segments. The flow rate changes in each segment depending on the location of the cutter. These cutting tables often have an added cooling bath beneath the plate (and the exhaust suction holes). To achieve good capture of the contaminants it is necessary to add an exhaust around the cutter on the upper side.
Many different types of tables exhaust downwards and have a perforated surface covering part of the table, e. g., inside a cabinet. This design is common in biological safety cabinets (see Section 10.4.6.4). Biological safety cabinets use a combination of a supply and an exhaust opening, and therefore are not defined as a downdraft table.
The downdraft tables used for small-scale chemical work are only recommended for small operations that generate contaminants near the working surface. The items that are used should not cover too large an area of the exhaust. Covering the surface increases the velocity and subsequently also the
Reach of the exhaust, but the contaminant generation usually takes place on the covered surface where velocities are too low for adequate capture, resulting in spread of contaminants to the room.
Use of warm processes on a downdraft table should be avoided since the air velocity created by the exhaust is often lower than the velocity due to buoyancy effects. Effective use of a downdraft table for welding requires velocities high enough to counteract the buoyancy, which could result in disturbances of the welding process.
The supply openings in the room must be placed to avoid disturbing the flow into the table. This naturally is easier to accomplish when the table is furnished with side walls.
Also, when using cold processes on a downdraft table, the worker should avoid leaning over the working place or sitting too close to it, since this could disturb the flow into the exhaust and increase the spread of contaminants into the workspace. The same effects as when using fume cupboards could easily be the result (Section 10.2.3.3).
An increase in flow rate could be advantageous since more contaminants could be captured. However, an increased flow could disturb the process and also generate drafts for the worker. A diminished flow rate results in less capture of contaminants.
When a built-in fan and filter is used and the air is recirculated into the workroom, it is necessary to consider the influence of the contaminants on the overall concentration inside the work room. Moreover, it could be difficult to dampen the noise from a built-in fan with a large flow rate. Small tables do not have these problems, since they are connected with ducts to a central fan.
Design Equations and/or Parameters
For small downdraft tables used for chemical work, a flow rate of 0.28 m3 s_! and m2 table is used. This gives a mean velocity immediately above the surface of approximately 0.3 m s’1. This is a very low velocity, which can not capture moving contaminants. When these values are used, a maximum use height of 0.15 m is recommended, which should result in a small leakage from the source to the surrounding. This presumes there is at least 0.1 m of uncovered surface between the worker and the source and that the surface is covered to less than 30%. For these tables the pressure difference is between 50 and 100 Pa, depending on the density of holes.63
The larger tables (0.8 x 0.5 m to 1.2 x 0.8 m) used for grinding, etc. normally operate with mean velocities of approximately 1 to 1.5 m s-1 at the surface. (There are prefabricated tables with face velocity equal to 0.5 m s-1). Mean velocities of 1 to 1.5 m s-1 result in flow rates of 3600 and 5400 m3 s-1 and m2. Some tables use velocities between 2 and 9 m s-1, resulting in flow rates of 7200 and 64 800 m3 s-1 and m2. These are very high flow rates, which result in noise and high costs. However, at a distance of 0.5 m above the table surface the velocity in the center is 3 m s’1. Tables with these velocities have a good capability to capture most generated particles. The velocity is calculated by using the equations for a square flanged opening 1 m2 in size and an opening velocity of 9 m s”1 (Section 10.2.2).
Recirculation is commonly used for grinding tables due to the large exhaust airflow rates of this design. For these larger tables the free area is much larger than for the small chemical tables, resulting in a much lower pressure difference. In order to maintain an even distribution of airflow over this large surface, it is necessary to use
Some type of flow distribution device. The grating bars are usually not enough, and filters placed immediately below the table could have a pressure difference large enough to distribute the flow rate over the entire surface (see Fig. 10.38).
For evaluation the velocity distribution and capture velocity could be used. Since the worker is quite close to the contaminant-generating place, occupational hygiene efficiency is possible (Section 10.5).
Posted in INDUSTRIAL VENTILATION DESIGN GUIDEBOOK