Design Manual

For Treatment of Process Water, Potable Water & Wastewater

Built By:

WaterSmart Environmental, Inc.

Shawnee Mission, Kansas

Featuring Inclined Corrugated Plates

WSE Publication No. 796

Particle/Liquid Separation Systems

and Tutorial

Prepared for: Website Visitor

WaterSmart Environmental, Inc. is a manufacturer of highly engineered water purification components and systems. The company designs and builds a wide variety of water treatment equipment including packaged water and wastewater treatment plants, UltraPaq™ aerobic package plants, OAT™ Process anaerobic digesters with associated energy production, aerators, filters, dissolved air flotation separators, air strippers, complete skid assembled aqueous waste treatment plants, FilterFresh™ skid mounted potable water production plants, skid mounted wastewater treatment systems for laundromats, laundries, and car/truck wash facilities with water reclamation and reuse, softeners, demineralizers, activated carbon treatment equipment, and water purifiers for domestic and international markets. The company provides laboratory testing services, treatment recommendation, equipment sizing and specifications, plan layouts, and budget price estimates. Its products are sold nationally and internationally through locally based sales representatives, distributors, or licensees. The Energy and Power Management Division of the company designs and provides Energy Management Control Systems for large office buildings, malls, hospitals, and similar facilities. Reduction of combined or total energy costs by 30% or more can frequently be achieved. WaterSmart Environmental, Inc (WSE), manufactured PuriSep™ separator plants are designed as modern, cost-effective particle removal equipment for treatment of process water, potable water, and wastewaters by municipal, industrial, and governmental users. In general, separators are used to remove large quantities of small particles from a liquid flow stream. These particles range in size from 1 to 250 microns in size with varying densities. For the vast majority of applications, separators are today regarded as better performing and more cost-effective particle removal devices than traditional clarifiers. The company also supplies RainDrain™ Perimeter Trench Dual Media Filtration Systems (see WSE Publication No. 2195) for stormwater treatment through its national marketing organization. In addition, the company has pioneered the development of the OPCT™ (Optimized Physical/Chemical Treatment, see WSE Publication No. 995) process which may be considered whenever chemical precipitation is one of the treatment options. PuriSep™ separators are classified according to their treatment application. Occasionally, more than on type of process treatment will be required to achieve the desired degree of particle removal. Accordingly, PuriSep™ separation systems may include flocculation, chemical feed, dissolved air flotation (DAF), and multiple stage treatment. The PuriSep™ technology permits the removal of any kind or type of particles from a liquid, regardless of its size, density, concentration, or presence of other particles. WaterSmart can recommend the most suitable treatment system for your particle separation project. Fundamental separator theory and general design concepts are covered under each product’s treatment application section. WaterSmart Environmental, Inc. takes great pride in supplying treatment equipment and en-ergy management control systems that work as represented and as required for each applica-tion undertaken.

Table of Contents

General Theory

Differential Gravity Separation Of Parti-cles 1

Reynolds Number 2 Design Differences 4 Effect Of Plate Spacing 4 Materials Of Construction 5 Froude Number 5 Internal Hydraulics 5 Particle Coalescence 6 PuriSep™ Separators 6 Application Engineering Considerations 7 Particle Removal By Simple Gravity 8 Particle Removal By Induced Gravity 8 Flocculation 10 Dissolved Air Flotation 11 DAF Air Dissolution 12 DAF Air Release 13 Summary 14

Table I Water Viscosities & Densities 15

Table II API & Baumé Gravity Tables & Weight Factors 16

Table III Viscosity & Specific Gravity Of Com-mon Liquids 17

Table IV Viscosity Conversion Table 18

Solids Separation Application Data 19

Table V Common Surface Loading Rates 20 Sludge Removal Considerations 20 Solids Separator Selection 20

Table VI Overflow/Settling Rates 21 Application Data Sheet 23

Oil Water Separators Application Data 25

Types Of Oily Impurities And Their Separation 26

Table VII Treatment Options 27 Emulsion Breaking 27 Sizing Procedure 27

Table VIII Sources Of Oily Waste 28True Inclined Plate Separators 28Plate Inclination Angle 32Free Oil Removal 33

Table IX Projected Separation Area 32Table X Particle Concentrations 33

The Relationship Between Projected And Effective Plate Separation Area 33Surface Area 34Square Feet Of Projected Area Per Cubic Foot Of Media 34

Table XI Plate Spacing Per Angle Of Inclination 35

The Relationship Between Capacity And Performance 35

Oil Particle Size To Be Removed 35 API’s STS Number 36

Treatment Of Mechanically Elmusified Oil 37

Treatment Of Chemically Elmusified Oil 37 Treatment Of Dissolved Oil 37

Upgrading Existing Inadequate Oil/Water Separators 37

Coalescers 39 Sludge Removal Considerations 39 Maintenance 39 Flow Bypassing 39 Fixed Oil Weirs 40 Marketplace Misrepresentation 42 Recommended Specifications 43 Laboratory Testing 43 Application Data Sheet 45

Flotation Separators Application Data 47Table XII PuriSep™ System By Types Of Parti-

cles Present 48 Sludge Removal Considerations 48 DAF Separator Selection 48 Laboratory Testing 48 Application Data Sheet 49

General Theory

Differential Gravity Separation of Particles Gravity clarification permits removal of particles that exhibit densities different from their carrier fluid. Sepa-ration is accomplished by detaining the flow stream for a sufficient time to permit particles to separate out. In water and wastewater treatment practice, particles which rise to the surface of a liquid are said to possess Rise Rates, while particles which settle to the bottom exhibit Settling Rates. Both types obey Stoke's Law, which establishes the theoretical terminal velocities of the rising and/or settling particles. Stoke's Law is:

Vp = 2)Dcdp(d18

G−

η

where Vp = rising or settling velocity of discrete particleG = gravity constant

η = absolute viscosity of carrier fluid

dp = density of particle to be removed

dc = density of carrier fluid

D = diameter of discrete particle

A negative velocity is referred to as a RISE RATE. Oil scum, fats, greases, waxes, and other such particles that float to the surface of a liquid are said to possess Rise Rates. A positive velocity is referred to as a SETTLING RATE. All particles that settle to the bot-tom of a carrier fluid are said to possess Settling Rates. More often than not, both rising and settling particles are present in the same fluid stream.

Traditional separation equipment frequently consists of circular and rectangular clarifiers. In a rectangular unit, for example, of height "H", surface area "A", and hydraulic capacity, or flowthrough, of "Q" (see Figure 1), the residence time "Tr " of the liquid can be determined according to the following equation:

Tr = A x H

Q

Now assume that a discrete particle with a positive set-tling velocity or rate is suspended at the inlet top of the clarifier at point "⊕“. The time it takes, designated as "Tp", for this particle to reach the bottom can be found

by dividing the height of the clarifier by the particle set-tling velocity, or

Tp =H

pV

Particle "⊕“ will reach the bottom of the clarifier (and thereby become separated from its carrier fluid) if the residence time of the liquid is at least equal to the set-tling time of the particle. This condition may be ex-pressed as:

Tr = Tp

In substituting for Tr and Tp we have

A x H

Q=

H

pV

which reduces to:

Vp =Q

A (Equation #1)

Figure 1

The ratio Q/A is designated as the surface overflow rate or the surface loading of the clarifier and is expressed in either gpm/sq ft or gpd/sq ft. The surface overflow rate, then, is but a shorthand method of expressing the separation velocity Vp of the particle, which the clarifier will remove.

Equation #1 is most significant to separator technology because it proves that neither the height “H” of the clarifier nor its residence time “Tr” influence particle removal efficiency. The relationship, rather, between flow and surface area (or settling area) is controlling.

The foregoing elementary analysis of particle behavior applies to the ideal settling of discrete particles. Actual

results on full sized equipment are influenced by density and thermal currents, surface wind conditions, design of inlet hydraulic distributors, design of outlet hydraulic collectors, solids loading, solids characteristics, chemical feed, and other factors.

The routine use of the word surface area in traditional clarifier practice is extremely unfortunate because most oil/water separator manufacturers intentionally overrate the capacity of their equipment by claiming the actual surface area of the device rather than its projected separation area. This ruse is disguised by the use of terms like effective coalescing area, coalescing area, and separation area knowing full well that a well-intentioned user will interpret the area representations in the traditional sense. For an additional discussion on this marketplace fraud see Surface Area, p. 34.

Reynolds Number The Reynolds Number, abbreviated Re Number, is a mathematical identification of the presence or absence of quiescent settling conditions. When the Re Number is less than 500, particles settle out of a flow stream according to the velocities determined by Stoke's Law. Flow is said to be laminar. Laminar flow conditions permit maximum particle separation to take place. A Re Number between 500 and 1000 indicates flow conditions which are neither laminar nor turbulent. Consequently, some settling does occur, but the efficiency of separation suffers in proportion to the increasing size of the Re Number. A Re Number of above 1000 indicates turbulent flow conditions under which settling becomes ever more hindered. The Re Number may be determined according to the expression:

Re =η

VDHD ××

Where HD = Hydraulic Diameter in feet

Note: When the fluid flow passageway has a cross section area other than a circle, the hydraulic diameter is then determined according to the relationship follow-ing:

HD =4 x cross sec tion area

wetted perimeter

D = Density of fluid in lbs/cu ft

V = Fluid flow velocity in ft/sec

η = Absolute viscosity in lbs/ft sec

In examining this equation, two things are readily seen. First, if one arbitrarily doubles the spacing between the plates, the Re Number is also doubled. Secondly, if one doubles the flow velocity between the plates, the Re Number is again doubled. Conversely, halving ei-ther the plate spacing or the flow velocity also halves the Re Number. This says that the optimum separator

will incorporate both minimized plate spacing and minimized flow velocity. And in order to minimize the flow velocity, the cross section area of the fluid flow path must be maximized.

The importance of these relationships cannot be over emphasized. Two separators with identical projected plate separation areas can exhibit dramatically different particle separation efficiencies due to separator design differences. As proof of this, lets determine the Re Number of the two separators, both using the same number of 4 foot by 8 foot plates inclined at 55° from the horizontal. By definition, then, both units contain the same projected plate separation area. Their re-spective efficiencies, however, as reflected by their respective Re Numbers, can vary considerably.

For the first separator, use a cross flow type (flow goes from side to side) separator with 3/4 inch plate spacing at a hydraulic loading of 0.5 gpm/sq ft with water @ 68°F as the carrier fluid. This is equivalent to a surface-loading rate of 720 GPD/sq ft

Re =η

V x D x HD

Step #1

HD =4x8 ft (3 / 4" 12"/ ft)

8 ft 8 ft

² ‚

+

= 0.125 ft Step #2 D = 62.254 lbs/cu ft @ 68°F

In order to determine the fluid flow velocity through the separator, the flow capacity must be determined. A single plate contains 8 ft x 4 ft or 32 sq ft of area. At the stated 55° angle of inclination, each plate has a projected plate separation area of 32 x the cosine of 55° or 18.35 sq ft

At the stated hydraulic loading of 0.5 gpm/sq ft, the capacity is found by multiplying the projected plate separation area by the hydraulic loading. This calculates out to 9.175 gpm. Since the plates are spaced at ¾ inch, the flow passageway cross section area may be determined by multiplying the plate length by the spacing or 8 ft x ¾ inch. Dividing this product by 12”/ft yields 0.5 sq ft. Dividing the fluid flow capacity by the cross section area determines the linear fluid flow velocity, or:

Step #3

V =9.175 GPM

60 sec / min x 7.481 gal / cu ft x 0.5 sq ft

= 0.0409 ft/sec Step #4 η = 0.00067533 lbs./ft-sec

The Re number may now be calculated as follows:

Re =0.125 x 62.254 x 0.0409

0.00067533

= 471 (laminar flow conditions)

Note: Tables of water viscosities and densities are pro-vided on pages 15 through 18.

For the second separator, use an upflow (flow goes from bottom of separator to top) type with ¾ inch plate spacing at a hydraulic loading of 0.5 gpm/sq ft with water @ 68°F. Again, the surface-loading rate is 720 GPD per square foot and the plates are 4 foot by 8 foot inclined at 55° from the horizontal. The single difference between the second separator and the first is the direction of flow through the unit.

Step #1

HD =ft 4 + ft 4

/ft)12" (3/4" x ft 4 x 4

= 0.125 ft

Step #2

D = 62.254 lbs/cu ft @ 68°F

The flow capacity again calculates out to 9.175 GPM but the cross section area of the flow passageway now becomes 0.25 sq ft

Step #3

V = 9.175 gpm

60 sec / min x 7.481 gal / cu ft x 0.25 sq ft

= 0.0818 ft/sec

Step # 4

η = 0.00067533 lbs/ft-sec

The Re Number may now be calculated as follows:

Re = 0.125 x 62.254 x 0.0818

0.00067533

= 943 (non-laminar flow conditions)

The efficiency of the cross flow type separator may be calculated according to Stoke’s Law because particle separation occurs under laminar flow conditions. In the second example, since hindered settling conditions exist (Re Number greater than 500), the significantly lesser efficiency must be determined experimentally. The surface-loading rate of the second separator would have to be reduced by a factor of two to equal the Re Number of the cross flow type separator.

A number of oil/water separator manufacturers insist that any Re Number of less than 2000, rather than 500, reflects laminar flow conditions. These manufacturers have yet to learn the fundamental difference between closed conduit and open channel hydraulics.

Separator Designs The horizontal plate separator as shown in Figure 2 was introduced in 1904. Initially, the horizontal plate separator would work very well. By adding intermediate plates, particles had less distance to traverse before becoming removed (separated) from the flow stream. And since the number of plates used now increased the surface area, more flow could be treated without a decrease in performance, even without an increase in equipment size! In addition, laminar flow hydraulic conditions (Reynolds Number of less than 500) were easy to achieve because of the now much smaller hy-draulic diameter provided by the relatively short dis-tance between the plates. Under laminar flow condi-tions, particles behave according to Stoke's Law, pre-viously discussed.

Figure 2

As separated particles began to build up and collect on the floors and ceilings of each plate, evidencing re-moval of both rising and settling particles, the flow path of the liquid would become partially restricted. As a result, laminar flow conditions gradually disappear be-cause of induced turbulence, and once separated par-ticles would then become reentrained in the flow stream. The horizontal plate separator would then have to be cleaned of separated particles in order to restore it to its prior efficiency.

In order to overcome the routine cleaning that was ne-cessitated by the horizontal plate configuration, the technology was significantly improved when plates were inclined at an angle greater than the angle of re-pose of the separated particles creating the inclined plate separator. This plate inclination improvement shown in Figure 3 permitted both rising and settling particles to be removed from a flow stream simultane-ously and continuously without periodic cleaning.

A separator with inclined plates exhibits self-flushing or self-cleaning characteristics if:

1. The plates are non-oleophilic (oleophilic is the property of attracting and holding oily particles)

2. The angle of plate inclination exceeds the angle of repose of the separated particles, and

3. The smoothness of the plates is sufficient to permit the separated particles, to slide down (or up) and out of the flow stream.

Figure 3 Adding parallel plates and then by installing them at an angle of inclination have significantly improved clarifier technology. Several separator manufacturers have used the above described separation principles to arrive at various designs. One supplier attaches a vibrator to the parallel plate assembly to encourage the separated particles to slide out of the flow stream rather than cling to the flat plates. Another directs the fluid flow from the top of the separator to the bottom in order to achieve maximum separation of the oil and scum particles. Yet others direct the flow from the bottom of the separator to the top in order to achieve maximum separation of those particles that exhibit positive settling rates (sediment removal). And some suppliers still use almost horizontal plates, like Hazen in 1904, that require periodic backflushing to clean the separator of deposited particles. Not to be outdone, other suppliers direct the inlet flow through the sides of the inclined plate separator where the liquid flow must turn 90 degrees to exit from the top of the device.

Tube settlers, whether circles, hexagons, diamonds, square, or chevrons in cross-section, are further ver-sions of the classical plate separator. The chief and continuing criticism of the tube settlers is their suscep-tibility to entrain small particles. As these small parti-cles are separated from the flow stream they slide out of the tube settler and directly into the incoming flow, thereby becoming reentrained. As this reentrainment phenomenon increases the concentration of these par-ticles eventually closes off the flow. The now plugged

tube settlers must be cleaned in order to restore them to useful particle separation service.

Design Differences In order to appreciate fully the differences between various separator designs, an understanding of the interrelationships between the following design consid-erations must be developed:

1. Reynolds Number 2. Internal separator hydraulics and separated

particles flow paths, and 3. Particle coalescence

Effect of Plate Spacing Plate spacing is the shortest distance between adjacent surfaces. When considered alone, plate spacing does not increase or decrease the Re Number when hydraulic loading rates are constant. As plate spacing is increased, hydraulic diameter is increased by an amount equal to the corresponding decrease in fluid flow velocity. Since both of these terms vary in direct proportion to the Re Number, they cancel each other out. The reverse is also true. If plate spacing is decreased, hydraulic diameter is also decreased but in the same proportion as the fluid flow velocity is increased. Consequently, plate spacing has no practical impact on the Re Number so long as the hydraulic loading, or surface overflow rate, remains a constant.

Plate spacing, however, does have a significant impact on the surface overflow rating of a separator. The wider the plate spacing, the further the distance a particle must traverse in order to separate from the fluid stream. As an example, a particle in a separator with 2 inch plate spacing must traverse 2.66 times more distance than the identical particle in a separator with ¾ inch plate spacing. In order for each of these separators to remove the same particle, the separator with 2 inch plate spacing would have to be rated at a surface overflow rate of only 37.6% that of the ¾ inch plate spaced separator.

This is calculated as follows. The vertical distance a particle must traverse is given by the plate spacing di-vided by the sine of (90° minus the angle of inclina-tion). In the case of a 55° inclination angle and two separators with ¾ inch and 2 inch plate spacing, the two settling distances are:

¾” ÷ sin (90° - 55°) = 1.31”

2” ÷ sin (90° - 55°) = 3.49”

3.49 ÷ 1.31 = 2.66

100% ÷ 2.66 = 37.6%

In summary:

Wider plate spacing can result in hindered settling conditions because of Re Numbers exceeding 500,

and always reduces the particle removal efficiency of separation, degrading effluent treatment quality.

Narrower plate spacing, however, favors laminar flow and quiescent settling conditions in accordance with Stoke’s Law, and always increases the particle re-moval efficiency of separation thus optimizing the ef-fluent treatment quality.

Materials of Construction Separator Plates The separator plate material of construction is ex-tremely critical as it strongly impacts on performance. It is of paramount importance that the plates are smooth on both sides to enhance continuing migration of sepa-rated particles and non-oleophilic to maintain non-sticky plate surface

Keeping the separator module free from separated particles, particularly sludge particles, decreases or eliminates the necessity to clean the plate surfaces. Smooth non-oleophilic plates represent a preferred material such as fiberglass. By comparison, PVC plates are far from smooth and polypropylene plates are oleophilic. Oleophilic plates adsorb oil on its sur-faces thereby preventing the migration of sludge parti-cles. The final result is a plugged module.

Vessel Fiberglass, epoxy painted steel, aluminum, stainless steel, and concrete are all acceptable materials of con-structions for the separator vessel. The principal con-sideration respecting materials relates to the intended service, i.e., depth of bury in the case of an under-ground installation, resistance to corrosion—both inte-rior and exterior surfaces, compatibility of surfaces in contact with the carrier fluid and any contaminants to be separated, and installation requirements. Should a proposed stormwater separator be located in a high groundwater location, a factory packaged metal or fi-berglass separator would install far easier than a pour-in-place concrete vessel. If the separator application is small enough, a precast concrete vessel may well be appropriate because of the ability to install quickly. Concrete separators also provide significant ballast because of their inherent weight many times eliminat-ing the need for a ballast pad. The preferred vessel material of construction, therefore, is not dependent upon the particle separation application under consid-eration.

Froude Number The Re Number is vital to separator design and per-formance. It is defined as:

ReNumberInertia force per unit area

Viscous force per unit area=

The Froude Number (Fr) is another number sometimes referenced in the literature of separator suppliers. It is defined as:

Fr NumberInertia force per unit area

Gravity force per unit area=

Basically, the Fr Number is the ratio of the velocity of flow to the celerity of a small wave in quiet fluid. Con-sequently, if the Fr Number is less than 1.0, a wave is swept downstream. The Fr Number is of importance only when there exists a sloping interface of two fluids of different density. It is significant in the study and behavior of flow in open channels. In the design and performance of separators, however, the Fr Number possesses no significance whatsoever. Some separa-tor manufacturers incorrectly reference the Fr Number in their sales literature.

Internal Hydraulics The internal hydraulics of a separator have a direct impact on the Re Number and consequently on particle separation conditions. At a specific hydraulic loading rate, the shortest flow path through the separator mod-ule will always result in the smallest flow velocity and therefore the lowest possible Re Number. To achieve a short flow path requires a large entry plane into the separator along with appropriate hydraulic flow distri-bution and subsequent treated water collection. The consequences of inadequate hydraulic distribution and collection are under-utilization of plate surface area. Likewise, separator designs with small entry planes possess inherently higher Re Numbers and thus must be rated at reduced hydraulic loadings.

If the path of the particle to be separated is opposite to, or countercurrent with, the path of the liquid flow, some of the particles with lesser settling/rise rates will be swept back into the liquid flow. This phenomenon is referred to as particle reentrainment. Particle reen-trainment also occurs whenever a migrating particle passes through the flow stream en route to its ultimate destination.

Particle reentrainment always degrades the quality of the treated water effluent as well as the efficiency of separator performance. Particle reentrainment may be effectively eliminated by:

1. Directing the paths of the separated particles and the liquid flow stream so that their respec-tive paths are 90° from each other. This ar-rangement can only be achieved in a cross flow type separator.

2. Causing the smaller particles (those discrete particles possessing lesser settling/rise rates) to coalesce together thereby agglomerating and forming larger particles. Particle settling/rise rates are proportional to the square of their di-ameters according to Stoke’s Law previously discussed. Particle coalescence therefore en-hances separator performance while minimizing particle reentrainment. Maximum coalescence is achieved when using corrugated plates.

Some oil/water separator manufacturers deploy two or more separator modules in series flow as shown in Figure 4.

In order to properly claim the projected area of an in-clined plate separator, the hydraulic continuum within the module must be continuous. Whenever a second or third module is used, the hydraulic continuum is in-terrupted thus restarting the particle separation proc-ess upon each module entry. In theory, additional modules in series flow impart very little additional parti-cle removal benefit.

Figure 4

Particle Coalescence Particle coalescence is important to enhance because a particle’s rate of separation, either rising or settling, is directly proportional to the square of its diameter. By inducing smaller particles to join together to form larger particles (called coalescence), the efficiency of particle separation is increased accordingly.

As separated particles migrate along a flat plate, some particle agglomeration takes place due to the collision between particles exhibiting different migration rates. This agglomeration takes place in the direction of parti-cle migration. As separated particles migrate along a corrugated plate, they slide to the trough of the corru-gations. This phenomenon is another example of coa-

lescence. As these coalesced particles continue mi-grating, they also agglomerate due to collisions be-tween particles exhibiting different migration rates. However, in a corrugated plate separator, particle ag-glomeration is achieved in two directions, namely, the direction of particle separation as well as the direction of separated particle migration. Thus, maximum parti-cle coalescence and agglomeration can only be achieved in a corrugated plate separator. One can therefore expect higher levels of separator perform-ance from a corrugated plate separator. Figure 5 illus-trates particle coalescence.

Figure 5

PuriSep Separators By combining the preferred design principles in a sin-gle separator, maximum particle removal efficiency and thus ideal process performance is achieved. The most modern, state-of-the-art, and optimum separator has been realized with the PuriSep corrugated plate separator manufactured by WaterSmart Environ-mental. See Figure 6.

The cross flow design permits simultaneous separation and self-flushing of both rising and settling particles without favoring one type of particle over the other. This dual-phase particle removal capability often per-mits single stage process treatment rather than multi-ple stage treatment. Dual phase particle removal is especially important when raw water characteristics change or vary through ranges of concentrations. The PuriSep has a wide entry plane and correspondingly short flow path thus minimizing the Re Number. Three distinct flow pressure drops assure maximum flow dis-tribution and consequently complete plate utilization. Additionally, since the directions of liquid and sepa-rated particle flow are always 90° to each other, parti-cle reentrainment is virtually eliminated. And by pro-viding an especially smooth finish on each side of the corrugated plates, particle coalescence is enhanced. No other separator design utilizes as many preferred performance features as the PuriSep cross flow in-clined corrugated plate separator.

TOP OF VESSEL

BOTTOM OF VESSEL

FLOW FLOW

The major advantages of the PuriSep separator are: 1. Minimum equipment space--less than any other

separator available. 2. Low Re Numbers, therefore quiescent settling

conditions and maximum particle removals. 3. Smooth corrugated plates for enhanced particle

coalescence that induce maximum particle ag-glomeration and self-flushing characteristics.

4. Cross flow hydraulic path that minimized both the Re Number and separated particle reen-trainment.

5. Dual phase particle separation, simultaneously, without favoring one type of particle over the other.

Figure 6

Application Engineering Considerations Projected Plate Separation Area To determine hydraulic loading rates of a separator, the flow and surface area are used in the same manner in which they are used to size, or rate, traditional clarifiers. The projected separation area of an inclined plate separator is determined by adding up the horizontally projected areas of the several plates as is illustrated in Figure 7.

The projected separation area of a separator inclined at 60° to the horizontal, for example, would be exactly 50% of the total inclined area since the cosine of 60° is 0.500. Since separators are not “perfect” separation devices, separator manufacturers multiply the total projected separation area by an “efficiency factor” which, when multiplied by the total projected separation area, produces the true separation area, or “effective separation area” of that particular unit. The efficiency factor will normally range from 0.95 in highly efficient separators to about 0.35 in highly inefficient

separators. The efficiency factor takes into account separator internal hydraulics, plate utilization, flow dampening, and separator module inlet and outlet flow disturbances.

Figure 7

Particle Removal By Simple Gravity The most fundamental type of gravity separation is the case where the flow stream contains particles that adequately settle out. Flow streams that contain particles with insufficient separation velocities to permit gravity settling will be considered later on.

The particles contained in this fundamental type of flow stream are frequently mixed. Some of the particles will exhibit rise rates and thus will eventually rise to the surface of the liquid. The balance of the particles may exhibit settling rates and thus will eventually settle to the bottom of the separator. In order to consider any clarifier or separator for simple gravity separation, sufficient separation velocity, expressed as either a rise rate or a settling rate (or both), must exist so that these particles separate out in a reasonable length of time. The size of a particle and the difference in specific gravity or density between the particle and the carrier fluid together determine the rate of separation characterized by that particle. Stoke’s Law gives the relationship between a particle and its separation velocity. Simple gravity separation may be accomplished with the separator shown in Figure 8.

Figure 8--Oil/Water/Solids Separator

Since all PuriSep separators simultaneously remove both rising and settling particles, the above separator may be used either as a Solids Separator or as an Oil/Water Separator. Even though the principal appli-cation may change, the equipment design remains es-sentially the same. If the raw water contains both ris-ing and settling particles, the above separator be-comes ideal because of its ability to remove both types of particles without favoring one type over the other. Additionally, if the process application ever changes from one kind of requirement to another, the same unit accomplishes both kinds of treatment. Traditional separators are usually designed to remove either rising or settling particles, but not both.

Particle Removal by “Induced” Gravity If the particles to be removed do not possess sufficient rise or settling rates to permit their removal by simple gravity separation, additional treatment is required. A particle can possess an inadequate separation velocity for two reasons:

1. Particle size is small, usually less than 5 microns. Chemical emulsions are typical particles in this size range. Particle density, or specific gravity, of these small particles does not influence their separation rates, and/or

2. Particle density, or specific gravity, of the particle is too close to that of the carrier fluid. The particle therefore neither rises nor settles at a sufficient velocity to permit gravity separation. The particle is said to possess “neutral gravity”. The particle size of the neutral gravity particles has little influence on their separation rates.

Since the force of gravity is a constant, the only way to separate neutral gravity particles from their carrier fluid is to “induce” an added density differential to the particle. By inducing this differential, separation rates are increased thereby permitting gravity separation.

In the case presented by small particles, the inducement of added density differential is provided by increasing the size of the particle. As Stoke’s Law shows, the separation velocity of a particle is directly proportional to the square of its diameter. By increasing the size of a particle, one also increases, sometimes dramatically, its rate of separation.

In order to accomplish particle growth, it is necessary to add one or more chemicals to the flow stream fol-lowed by flocculation of the particles. This is graphi-cally illustrated in Figure 9.

Figure 9 This type of application is effectively accomplished with the combination flocculator/separator shown in Figure 10. The resulting agglomerated large particle shown above may exhibit either a rise or a settling rate. The direction of separation is not significant. Its rate of separation rather is the parameter that permits gravity separation.

Figure 10--Combination Flocculator/Separator

If the now agglomerated particle still exhibits a yet insufficient separation rate, as is entirely possible, further process treatment will be required in the form of Dissolved Air Flotation (DAF). In this induced density differential refinement, air saturated recycle water under elevated pressure is distributed in front of the separator. Excess air “boils” out of the recycle water on release of system pressure. These micron-sized air bubbles, or particles, then become attached or occluded within the agglomerated floc structure causing these neutral gravity particles to rise quickly to the liquid surface. This is graphically illustrated in Figure 11.

Figure 11 This type of application is effectively accomplished with the combination flocculator/DAF separator shown in Figure 12. Note that this separator is identical to the separator shown in Figure 10 except for the addition of the DAF distributor.

In order to remove small particles effectively, the first step is always chemical feed and flocculation. The resulting agglomerated particle may then exhibit a sufficient separation rate. If it does not, then DAF may be added to assist particle removal.

In referring back to case #2 where the particle density is too close to that of the carrier fluid, and thus exhibits neutral gravity, the removal technique is to add the DAF distributor once again. The now attached air permits rapid and efficient removal of these particles. This is graphically illustrated in Figure 13.

Figure 12--Combination Flocculator/DAF Separator

This type of application is effectively accomplished with the DAF separator shown in Figure 14. Note that this separator is identical to the separator shown in Figure 8 except for the addition of the DAF distributor.

Figure 13

In the examples cited, particles with adequate separation velocities, as well as small and otherwise neutral gravity particles, have all been removed from their carrier fluid. Those separation techniques may not work when several kinds of particles are simultaneously present, in significant quantities, in the same flow stream with total particle removal required.

As an example, a flow stream may contain

1. Particles with a sufficient rise rate to permit gravity separation and large neutral gravity particles with an insufficient separation rate, or

2. Particles with a sufficient settling rate to permit gravity separation and large neutral gravity particles with an insufficient separation rate, or

3. Both rising and settling particles along with neutral gravity particles.

Figure 14--DAF Separator

As has been shown, in order to remove the neutral gravity particles, DAF must be employed. However, in the presence of these other particles, DAF treatment would prove ineffective. In the case of the settling particles, the DAF “induced gravity differential” would change these into neutral gravity particles, and furthermore, these particles would dilute the concentration of air particles available to lift the neutral gravity particles to the surface. In the case of the rising particles, the dilution of air particles would occur causing insufficient air to permit removal of the neutral gravity particles.

Figure 15--Combination Gravity Separator/ DAF

Separator

To achieve effective particle removal, it is obvious that both rising and settling particles, to the extent present, must be substantially removed prior to the use of DAF. Indeed, the greatest single reason that DAF systems do not perform properly is the failure to recognize this requirement. A two stage separation system consisting of a first stage gravity separator followed by a DAF separator is necessary. This type of application is effectively accomplished with the combination gravity separator/DAF separator shown in Figure 15. Note that this separator is identical to the separator shown in

Figure 12 except that the first stage flocculator has been replaced by a first stage gravity separator.

One last example illustrates the necessity for sequential removal of specific kinds of particles in order to achieve effective treatment. In the case of a flow stream containing significant quantities of free oil (exhibiting an adequate separation velocity) along with significant quantities of chemically emulsified oil (exhibiting inadequate separation velocity), a three-stage separation system is required. The first stage of treatment would consist of a simple gravity separator to remove those particles with a sufficient separation velocity, namely, the free oil constituent of the contaminant. The remaining emulsified oil must first be chemically treated to “break” the emulsion followed by flocculation to permit particle agglomeration or particle growth. The second stage treatment would then consist of flocculation with chemical pre-treatment. Since the agglomerated particles would still exhibit neutral gravity because of their close proximity in density to their carrier fluid, the use of DAF would be required. The third stage of treatment then consists of a DAF separator.

The separation of particles from a liquid stream is not difficult so long as the above design principles are followed.

Determining raw water characteristics accurately is an obvious first step. Deciding upon level(s) of treatment is the second step. Thereafter, the design of an appropriate treatment system is relatively easy. Comprehensive laboratory testing may well be necessary. Occasionally, job site pilot plant testing may be advisable before full-scale equipment designs are considered. So long as the entire range of ancillary equipment and options are available to the treatment plant designer, the effective separation of virtually any particle, or any combination of particles, is achievable without resorting to complex treatment schemes. And since the PuriSep separation system utilizes any combination of flocculation, chemical feed, gravity separation, and DAF, its application may be relied upon as the exclusive treatment equipment system with single source responsibility provided by WaterSmart Environmental, Inc.

Flocculation Flocculators are used to enhance particle growth or particle agglomeration and floc formation after chemical addition. Flocculation always occur in two successive steps or stages:

1. Perikinetic Flocculation is particle growth or aggregation resulting from random thermal motion of fluid molecules. This random motion is known as the “Brownian Movement”. Coagulation and perikinetic flocculation take place on a micro scale in less than one (1) second.

2. Orthokinetic Flocculation is particle growth or aggregation resulting from induced velocity gradients in the fluid. Orthokinetic flocculation occurs slowly in flocculators, usually requiring several seconds or minutes for complete floc formation. This type of flocculation is the predominant mechanism in potable, process, and wastewater treatment.

Chemical feed dosage is generally determined by a laboratory procedure called Jar Tests. This method uses a series of beakers fitted with adjustable, slow RPM stirrers. Various dosages of chemicals, polymers, and other flocculents are added while stirring in order to achieve optimum floc producing conditions. The normal stirring time in this procedure is five (5) minutes. Traditional flocculators employ paddle mixers and baffles in continuously stirred tank reactors. In these types of flocculators, short-circuiting and back mixing always occur, simultaneously. The combination of variations in residence time coupled with widely varying velocity gradients (G values) is responsible for an entire range of resulting floc structures, sizes and corresponding settling (or rising) properties. As a consequence, flocculation is inefficient, requiring detention periods of from fifteen (15) to twenty five (25) minutes, or three to five times greater than the flocculation which was accomplished in the Jar Tests!

WaterSmart Environmental uses a modern corrugated plate hydraulic type flocculator, called CorruFloc, to achieve the same or better results as those determined in the Jar Tests. In the flocculator, chemically dosed water is directed through a series of compartments formed by adjacent corrugated plates that are assembled in a flocculator module. The velocity gradients within the flocculator may be adjusted by shifting every second corrugated plate.

Figure 17 In Figure 16, a minimum velocity gradient is generated entirely by the drag (friction due to liquid viscosity) between the plates. In Figure 17, a maximum G value has been established by a complete phase shift between the plates. In passing between the plates, the

water alternately speeds up and then slows down resulting in gentle, but positive, particle agglomeration or floc growth.

In a CorruFloc hydraulic flocculator, some particle separation takes place because floc formation is rapid. Particles that do separate either rise into a scum collection compartment where they are mechanically skimmed from the surface, and/or settle to the sludge collection compartment for subsequent transfer.

If the flocculation characteristics of the incoming water ever change, new phase shifts can be established between the corrugated plates that will once again result in optimum flocculation conditions. No other conventional flocculator can be adjusted over as wide a range of operating conditions as the CorruFloc flocculator. Velocity gradient changes of twenty to one (20:1) are feasible. This capability permits a wide accommodation of influent water characteristics. Process advantages are:

1. No moving parts thereby minimizing maintenance.

2. Short-circuiting and back mixing are eliminated. 3. Highly controlled velocity gradients assure

uniform and optimum floc formation. 4. Generous flocculation time resulting in

maximum particle growth on a consistent basis. 5. No mechanically induced shear forces to break

up floc growth (chief criticism of mechanical flocculators).

6. Maximum possible performance and reliability.

Dissolved Air Flotation Dissolved Air Flotation (DAF) is the process that utilizes small gas bubbles, in the size range from 3 micron to about 20 micron, to separate 5-10 micron neutral gravity particles from a carrier fluid. By neutral gravity is meant those particles that exhibit separation velocities (either rising or settling) of less than 1 inch per minute. The presence of other particles with separation rates greater than 1 inch per minute will tend to retard the efficiency of separation of neutral gravity particles in proportion to their concentration. Should these other particles be present in a concentration higher than 2,000 to 5,000 mg/L, depending on separation velocities, they must be removed in a gravity type separator prior to the DAF separation system in order to achieve the removal efficiencies inherent to the DAF process.

Practitioners who have failed to take the above limitations into account have frequently misapplied the Dissolved Air Flotation process. A clear understanding of the DAF process is necessary before it can be applied to a specific treatment application.

Induced Air Flotation is another “air assist” process in which air or gas is occluded within or injected into a liquid flow and thereafter permitted to contact neutral

Figure 16

gravity particles. Induced Air Flotation (IAF) bubbles are many times larger in size (roughly 100 times) than DAF bubbles. As a practical matter, for every one IAF bubble there are one million DAF bubbles. Therefore, the opportunity to make contact, based on random probability, with neutral gravity particles is proportionally much greater with the DAF process. Particle removal efficiencies are consequently far better with the DAF process.

In the DAF process, air (or other gasses) is mixed with water at elevated pressures as it is forced into solution. Excess air is bled off to prevent reentrainment into the DAF distribution piping. In order to illustrate the amount of air that can be dissolved in water under elevated pressures, Henry’s Law states that, for gasses of low solubility, the volume dissolved in water varies with absolute pressure. For example, at sea level and at a temperature of 20° C (68° F), water will dissolve approximately 2% of air by volume, but at 15 psig, the solubility is 4%, at 30 psig--6%, at 45 psig--8%, and at 60 psig--10%.

Upon a reduction in pressure, the water becomes su-persaturated with air causing the supersaturated por-tion to precipitate (boil) out in the form of micron sized bubbles. The amount of air that precipitates out is de-pendent upon the amount of saturation. The rate of air precipitation is dependent upon two factors. First is the degree of supersaturation, that is, the greater the de-gree the faster the rate. Second is the degree of turbu-lence after pressure release that the supersaturated water undergoes, that is, that the greater the turbu-lence the faster the rate.

These two factors are important considerations in the design of a DAF system because the ability of the process to remove neutral gravity particles is directly related to the release of air bubbles. The greater the concentration of neutral gravity particles to be removed, the greater the need for dissolved air bubbles to assist particle separation. Furthermore, the more efficient the air release phenomenon, the smaller, and less expensive, the DAF generation and distribution system.

DAF Air Dissolution There are four methods used to dissolve air in the application of the DAF process.

In one process the entire flow stream to be treated is first mixed with air under elevated pressure. System pressure is thereafter reduced to gravity pressure, which, in turn, permits the supersaturated portion of the air to boil out of solution in a reaction or separation vessel. Since the entire flow stream is pressurized, the pressurization system must be sized to accommodate the complete full flow stream. The system components include:

1. Pressure pump & motor 2. Air addition system 3. Pressure detention/reaction tanks 4. Pressure release valve 5. Interconnecting piping

In another method only part of the flow stream to be treated is mixed with air under pressure. The air saturated side stream (split stream) is then passed through a pressure release valve and discharges into the reaction/separation vessel. This method requires the use of a somewhat smaller DAF make-up system than the system first described in above.

A third technique uses pressurized treated water but recycles it into the raw water feed line where the entire flow must pass through a pressure release valve before discharge into the reaction/separation vessel.

The fourth method uses treated water for its DAF make-up system (called a recycle flow system), and thereafter discharges the air saturated water through a pressure release valve and into the reaction/separation vessel. This is the method used by WaterSmart Environmental, Inc.

In any DAF system, only the supersaturated portion of the dissolved air can be released. If the incoming raw water has little or no dissolved air, as is often the case, sufficient air must be dissolved to saturate this entire flow stream at gravity or ambient pressure conditions. The amount of air required to accomplish this initial degree of air saturation remains in solution and is consequently unavailable to accomplish air flotation. It therefore becomes desirable to minimize the amount of wasted dissolved air in any DAF system. Figure 18 on page 13 shows flow schematics of typical DAF separation systems.

DAF systems may be further improved by using treated water in a recycle mode of operation. Rather than having little or no dissolved air, as is the case with raw water pressurization systems, the recycle water already contains some dissolved air in rough proportion to the recycle rate. This preexisting dissolved air further reduces DAF capital and operating expenses.

As will be further discussed under the Flotation Separators Application Data Section, the ability of a DAF system to remove particles is directly related to the gross amount of air released in the form of micron sized bubbles. The amount of air released is dependent upon the recycle rate and system pressure (plus other factors considered shortly). By determining the air release requirement, and in considering the raw water dissolved air, liquid temperature, flow variations, and similar matters, one may design a more cost-effective DAF system for a particular treatment requirement.

Figure 18

DAF Air Release Supersaturated air begins coming out of solution within a few seconds after pressure release takes place. The supersaturated portion can continue coming out of solution for several minutes after pressure release has taken place. By this time, significant quantities of treated water, including the intermixed DAF water, will have been discharged from the reaction/detention vessel. To the extent that the discharged water contains still supersaturated air, the DAF system is

inefficient. It therefore becomes desirable to provide conditions that will accelerate the air release process so that it occurs within the reaction/detention vessel, and prior to its discharge.

One way to speed up the rate of air release is to design and operate the DAF water at higher rather than lower pressures. As stated prior, the rate of release is proportional to the difference between system pressure and released pressure. Additionally, the amount of air required to saturate the DAF water

(and thus wasted) in proportion to the supersaturated amount is less as the system pressure is increased. Higher system pressures, however, are more expensive to install and operate than lower system pressures. In balancing these competing cost considerations, it has been determined that a DAF system pressure between 50 and 75 psig is the most economical range for this type of treatment.

Another method of speeding up the air release rate is to provide high degrees of turbulence in the DAF water distribution system downstream from the pressure release valve. The first method to increase turbulence, and thereby the rate of air bubble discharge, is to maximize turbulence in the air release valve itself. Most DAF suppliers use a diaphragm type valve to accomplish air release. These valves work very well as pressure regulating control devices. The valve manufacturers intentionally hold the amount of turbulence created in these valves to a minimum. The interior of these valves operate in such a manner so as to create a converging section, a throat or flow control section, and lastly, a diverging section. In many respects, the valves simulate the hydraulics in a venturi meter. Flow indeed is controlled, as is its corresponding pressure. However, the amount and degree of turbulence is and has been minimized by its geometric design. In other words, diaphragm type pressure release or pressure relief valves do not impart a high degree of turbulence to their controlled flow. However, in the PuriSep DAF generation modules, a specially manufactured valve series is utilized which maximized hydraulic turbulence, and thus air bubble release, while also providing flow control and pressure release.

Most DAF suppliers discharge DAF water into the reac-tion/detention vessel directly from the exit of the pres-sure release valve. The DAF water then passes through but a single turbulence creating condition. In a PuriSep DAF separator, the DAF water is passed through a second turbulence creating condition con-sisting of a DAF distribution header containing flow distribution orifices. A modest, but important, amount of pressure drop is designed into the DAF distribution header in order to create a secondary turbulent condi-tion. In discharging the DAF water through two (2) tur-

bulence zones, air release (precipitation) efficiencies are maximized. Additionally, the DAF distribution sys-tems in the PuriSep DAF separators are located near the vessel bottom thereby assuring maximum vertical use of the DAF gas bubbles as they slowly rise to the surface of the separator vessel.

That the DAF process has frequently been misapplied cannot be argued. That the DAF air dissolution and air release systems in general use today have been poorly designed cannot be denied. That the DAF process remains a preferred treatment system, in spite of its somewhat high installation and operating costs is generous testimony to its process performance. By examining the various considerations discussed in this section, one may enhance the application and use of the DAF process. And by selecting a PuriSep DAF separator, process efficiencies and cost-effectiveness are both assured.

Summary WaterSmart Environmental, Inc. manufactures and services advanced water purification equipment for municipal, governmental, and industrial markets on a global basis. The PuriSep separation systems represent but a single product line. The engineering policy of the company is to consistently strive to:

Minimize Energy input. Equipment size and weight. Mechanical components. Treatment and transfer complexities.

While Emphasizing Conservative process and mechanical designs. Corrosion protection. Fail-safe features. OSHA compliance.

In Order to Achieve Minimum operation and maintenance costs. Optimum mechanical and electrical reliability. Performance requirements with room to spare. Cost-effective solutions on a systems basis.

Table I Water Viscosities & Densities

Temperature Absolute Viscosity of Pure Water

Density of Pure Water in Air

°C °F poises lbs/ft-sec gm/cc lbs/cu ft 0 32.0 0.017870 0.00120424 0.999 62.351 1 33.8 0.017280 0.00116338 0.999 62.355 2 35.6 0.016710 0.00112407 0.999 62.358 3 37.4 0.016180 0.00108799 0.999 62.360 4 39.2 0.015670 0.00105324 1.000 62.360 5 41.0 0.015190 0.00102059 0.999 62.360 6 42.8 0.014720 0.00098968 0.999 62.359 7 44.6 0.014280 0.00095984 0.999 62.357 8 46.4 0.013860 0.00093135 0.999 62.354 9 48.2 0.013460 0.00090460 0.999 62.350 10 50.0 0.013070 0.00087873 0.999 62.345 11 51.8 0.012710 0.00085427 0.999 62.339 12 53.6 0.012350 0.00084870 0.999 62.333 13 55.4 0.012020 0.00080824 0.999 62.326 14 57.2 0.011690 0.00078681 0.999 62.317 15 59.0 0.011390 0.00076631 0.999 62.309 16 60.8 0.011090 0.00074662 0.999 62.299 17 62.6 0.010810 0.00072761 0.999 62.289 18 64.4 0.010530 0.00070953 0.999 62.278 19 66.2 0.010270 0.00069206 0.999 62.266 20 68.0 0.010020 0.00067533 0.998 62.254 21 69.8 0.009779 0.00065920 0.998 62.241 22 71.6 0.009548 0.00064368 0.998 62.228 23 73.4 0.009325 0.00062883 0.998 62.213 24 75.2 0.009111 0.00061431 0.997 62.198 25 77.0 0.008904 0.00060054 0.997 62.183 26 78.8 0.008705 0.00058710 0.997 62.167 27 80.6 0.008513 0.00057420 0.997 62.150 28 82.4 0.008327 0.00056177 0.996 62.133 29 84.2 0.008148 0.00054967 0.996 62.115 30 86.0 0.007975 0.00053805 0.996 62.097 31 87.8 0.007808 0.00052682 0.995 62.078 32 89.6 0.007647 0.00051600 0.995 62.058 33 91.4 0.007491 0.00050552 0.995 62.038 34 93.2 0.007340 0.00049531 0.995 62.018 35 95.0 0.007194 0.00048550 0.994 61.996 36 96.8 0.007052 0.00047609 0.994 61.974 37 98.6 0.006915 0.00046682 0.994 61.952

Temperature Absolute Viscosity of Pure Water

Density of Pure Water in Air

°C °F poises lbs/ft-sec gm/cc lbs/cu ft 38 100.4 0.006783 0.00045788 0.993 61.929 39 102.2 0.006654 0.00044921 0.993 61.907 40 104.0 0.006529 0.00044081 0.992 61.885 41 105.8 0.006408 0.00043268 0.992 61.859 42 107.6 0.006291 0.00042475 0.992 61.833 43 109.4 0.006178 0.00041709 0.991 61.807 44 111.2 0.006067 0.00040970 0.991 61.781 45 113.0 0.005960 0.00040238 0.990 61.756 46 114.8 0.005856 0.00039532 0.990 61.729 47 116.6 0.005755 0.00038853 0.989 61.701 48 118.4 0.005656 0.00038188 0.989 61.674 49 120.2 0.005561 0.00037550 0.988 61.647 50 122.0 0.005468 0.00037138 0.988 61.620 51 123.8 0.005378 0.00036726 0.988 61.591 52 125.6 0.005290 0.00036313 0.987 61.562 53 127.4 0.005204 0.00035715 0.987 61.533 54 129.2 0.005121 0.00035137 0.986 61.504 55 131.0 0.005040 0.00034580 0.986 61.474 56 132.8 0.004961 0.00033908 0.985 61.443 57 134.6 0.004884 0.00033498 0.985 61.412 58 136.4 0.004809 0.00032470 0.984 61.381 59 138.2 0.004736 0.00031979 0.984 61.350 60 140.0 0.004665 0.00031502 0.983 61.319 61 141.8 0.004596 0.00031042 0.993 61.286 62 143.6 0.004528 0.00030575 0.982 61.252 63 145.4 0.004462 0.00030124 0.982 61.220 64 147.2 0.004398 0.00029688 0.981 61.186 65 149.0 0.004335 0.00029264 0.981 61.153 66 150.8 0.004273 0.00028848 0.980 61.119 67 152.6 0.004213 0.00028444 0.980 61.084 68 154.4 0.004155 0.00028048 0.979 61.050 69 156.2 0.004098 0.00027665 0.979 61.015 70 158.0 0.004042 0.00027289 0.978 60.981 71 159.8 0.003987 0.00026919 0.977 60.944 72 161.6 0.003934 0.00026556 0.976 60.908 73 163.4 0.003882 0.00026207 0.976 60.871 74 165.2 0.003831 0.00025864 0.975 60.835 75 167.0 0.003781 0.00025528 0.975 60.798

15

Table II API and Baumé Gravity Tables and Weight Factors

A.P.I Gravity

Baumé Gravity

Specific Gravity

Lbs Per US Gal

US Gals Per lb

A.P.I

Gravity Baumé Gravity

Specific Gravity

Lbs Per US Gal

US Gals Per lb

0 10.247 1.0760 8.962 0.1116 51 50.57 0.7753 6.455 0.1549 1 9.223 1.0679 8.895 0.1124 52 51.55 0.7711 6.420 0.1558 2 8.198 1.0599 8.828 0.1133 53 52.54 0.7669 6.385 0.1566 3 7.173 1.0520 8.762 0.1141 54 53.53 0.7628 6.350 0.1575 4 6.148 1.0443 8.698 0.1150 55 54.52 0.7587 6.316 0.1583 5 5.124 1.0366 8.634 0.1158 56 55.51 0.7547 6.283 0.1592 6 4.099 1.0291 8.571 0.1167 57 56.50 0.7507 6.249 0.1600 7 3.074 1.0217 8.509 0.1175 58 57.49 0.7467 6.216 0.1609 8 2.049 1.0143 8.448 0.1184 59 58.48 0.7428 6.184 0.1617 9 1.025 1.0071 8.388 0.1192 60 59.47 0.7389 6.151 0.1626 10 10.00 1.0000 8.328 0.1201 61 60.46 0.7351 6.119 0.1634 11 10.99 0.9930 8.270 0.1209 62 61.45 0.7313 6.087 0.1643 12 11.98 0.9861 8.212 0.1218 63 62.44 0.7275 6.056 0.1651 13 12.97 0.9792 8.155 0.1226 64 63.43 0.7238 6.025 0.1660 14 13.96 0.9725 8.099 0.1235 65 64.42 0.7201 5.994 0.1668 15 14.95 9.9659 8.144 0.1243 66 65.41 0.7165 5.964 0.1677 16 15.94 0.9593 7.989 0.1252 67 66.40 0.7128 5.934 0.1685 17 16.93 0.9529 7.935 0.1260 68 67.39 0.7093 5.904 0.1694 18 17.92 0.9465 7.882 0.1269 69 68.37 0.7057 5.874 0.1702 19 18.90 0.9402 7.830 0.1277 70 69.36 0.7022 5.845 0.1711 20 19.89 0.9340 7.778 0.1286 71 70.35 0.6988 5.817 0.1719 21 20.88 0.9279 7.727 0.1294 72 71.34 0.6953 5.788 0.1728 22 21.87 0.9218 7.676 0.1303 73 72.33 0.6919 5.759 0.1736 23 22.86 0.9159 7.627 0.1311 74 73.32 0.6886 5.731 0.1745 24 23.85 0.9100 7.578 0.1320 75 74.31 0.6852 5.703 0.1753 25 24.84 0.9042 7.529 0.1328 76 75.30 0.6819 5.676 0.1762 26 25.83 0.8984 7.481 0.1337 77 76.29 0.6787 5.649 0.1770 27 26.82 0.8927 7.434 0.1345 78 77.28 0.6754 5.622 0.1779 28 27.81 0.8871 7.387 0.1354 79 78.27 0.6722 5.595 0.1787 29 28.80 0.8816 7.341 0.1362 80 79.26 0.6690 5.568 0.1796 30 29.79 0.8762 7.296 0.1371 81 80.25 0.6659 5.542 0.1804 31 30.78 0.8708 7.251 0.1379 82 81.24 0.6628 5.516 0.1813 32 31.77 0.8654 7.206 0.1388 83 82.23 0.6597 5.491 0.1821 33 32.76 0.8602 7.163 0.1396 84 83.22 0.6566 5.465 0.1830 34 33.75 0.8550 7.119 0.1405 85 84.20 0.6536 5.440 0.1838 35 34.73 0.8498 7.076 0.1413 86 85.19 0.6506 5.415 0.1847 36 35.72 0.8448 7.034 0.1422 87 86.18 0.6476 5.390 0.1855 37 36.71 0.8398 6.993 0.1430 88 87.17 0.6446 5.365 0.1864 38 37.70 0.8348 6.951 0.1439 89 88.16 0.6417 5.341 0.1872 39 38.69 0.8299 6.910 0.1447 90 89.15 0.6388 5.316 0.1881 40 39.68 0.8251 6.870 0.1456 91 90.14 0.6360 5.293 0.1889 41 40.67 0.8203 6.830 0.1464 92 91.13 0.6331 5.269 0.1898 42 41.66 0.8155 6.790 0.1473 93 92.12 0.6303 5.246 0.1906 43 42.65 0.8109 6.752 0.1481 94 93.11 0.6275 5.222 0.1915 44 43.64 0.8063 6.713 0.1490 95 94.10 0.6247 5.199 0.1924 45 44.63 0.8017 6.675 0.1498 96 95.09 0.6220 5.176 0.1932 46 45.62 0.7972 6.637 0.1507 97 96.08 0.6193 5.154 0.1940 47 50.61 0.7927 6.600 0.1515 98 97.07 0.6166 5.131 0.1949 48 50.60 0.7883 6.563 0.1524 99 98.06 0.6139 5.109 0.1957 49 50.59 0.7839 6.526 0.1532 100 99.05 0.6112 5.086 0.1966 50 50.58 0.7796 6.490 0.1541

The above tables are based on the weight of 1 gallon (U.S.) of oil with a volume of 231 cubic inches at 60° Fahrenheit in air at 760 mm pressure and 50% humidity. Assumed weight of 1 gallon of water at 60° Fahrenheit in air is 8.32828 pounds.

The relation of Degrees Baumé or API to Specific Gravity is expressed by the following formulas:

For liquids lighter than water:

Degrees Baumé = 140

G130− ,

Baume Degrees+130140

=G

Degrees API = 141.5

G131.5− , G =

141.5

131.5 + Degrees A.P.I.

For liquids heavier than water:

Degrees Baumé = 145 - 145

G, G =

145

145 - Degrees Baume

G = Specific Gravity = ratio of the weight of a given volume of oil at 60° Fahrenheit to the weight of the same volume of water at 60° Fahrenheit.

Table III Viscosity and Specific Gravity of Common Liquids

Viscosity S.U.S. Liquid Specific Gravity 40° F 60° F 80° F 100° F 120° F 140° F 160° F

Miscellaneous Liquids Water 1.0 31.5 31.5 31.5 31.5 31.5 31.5 31.5 Gasoline .68-.74 30 30 30 30 30 30 30 Jet Fuel .74-.85 35 35 35 35 35 35 35 Kerosene .78-.82 42 38 34 33 31 30 30 Turpentine .86-.87 34 33 32.8 32.6 32.4 32 32 Varnish Spar .9 3500 1600 1000 650 530 250 230 Fuel Oil and Diesel Oil #1 Fuel Oil .82-.95 40 38 35 33 31 30 30 #2 Fuel Oil .82-.95 70 50 45 40 #3 Fuel Oil .82-.95 90 68 53 45 40 #5A Fuel Oil .82-.95 1000 400 200 100 75 60 40 #5B Fuel Oil .82-.95 1300 600 490 400 330 290 240 #6 Fuel Oil .82-.95 70000 20000 9000 1900 900 500 #2D Diesel Fuel Oil .82-.95 100 68 53 45 40 36 35 #3D Diesel Fuel Oil .82-.95 200 120 80 60 50 44 40 #4D Diesel Fuel Oil .82-.95 1600 600 280 140 90 68 54 #5D Diesel Fuel Oil .82-.95 15000 5000 2000 900 400 260 160 Crankcase Oils - Automobile Lubricating Oils SAE 10 .88-.935 1500-2400 600-900 300-400 170-220 110-130 75-90 60-65 SAE 20 .88-.935 2400-9000 900-3000 400-1100 220-550 130-280 90-170 65-110 SAE 30 .88-.935 9000-14000 3000-4400 1100-1800 550-800 280-400 170-240 110-150 SAE 40 .88-.935 14000-19000 4400-6000 1800-2400 800-1100 400-550 240-320 150-200 SAE 50 .88-.935 19000-45000 6000-10000 2400-4000 1100-1800 550-850 320-480 200-280 SAE 60 .88-.935 45000-60000 10000-17000 4000-6000 1800-2500 850-1200 480-580 280-380 SAE 70 .88-.935 60000-120000 17000-45000 6000-10000 2500-4000 1200-1800 580-900 380-500 Transmission Oils - Transmission Gear Lubricants SAE 90 .88-.935 14000 5500 2200 1100 650 380 240 SAE 140 .88-.935 35000 12000 5000 2200 1200 650 400 SAE 250 .88-.935 160000 50000 18000 7000 3300 1700 1000 Other Oils Castor Oil .96 36000 9000 3000 1400 900 400 300 Chinawood .943 4000 1800 1000 580 400 300 200 Coconut .925 1500 500 250 140 100 70 60 Cod .928 1800 600 300 175 110 80 70 Corn .924 1600 700 400 250 175 100 80 Cotton Seed .88-.925 15800 600 300 176 125 80 70 Cylinder .82-.95 60000 14000 6000 2700 1400 1000 400 Navy No. 1 Fuel Oil .989 4000 1100 600 380 200 170 90 Navy No. 2 Fuel Oil 1.0 24000 8700 3500 1500 900 480 Gas .887 180 90 60 50 45 Insulating 350 150 90 65 50 45 40 Lard .912-.925 1100 600 380 287 180 140 90 Linseed .925-.939 1500 500 250 143 110 85 70 Raw Menhaden .933 1500 500 350 140 110 80 70 Neats Foot .917 1000 430 230 160 100 80 Olive .912-.918 1500 550 320 200 150 100 80 Palm .924 1700 700 380 221 160 120 90 Peanut .920 1200 500 300 195 150 100 80 Quenching 2400 900 450 250 180 130 90 Rape Seed .919 2400 900 450 250 180 130 90 Rosin .980 28000 7800 3200 1500 900 500 300 Rosin (Wood) 1.09 Extremely Viscous Sesame .923 1100 500 290 184 130 90 60 Soy Bean .927-.98 1200 475 270 165 120 80 70 Sperm .883 360 250 170 110 90 70 60 Turbine (Light) .91 500 350 230 150 Turbine (Heavy) .91 3000 1400 700 330 200 150 100 Whale .925 900 450 275 170 140 100 80

Table IV Viscosity Conversion Table

SUS Saybolt Universal Seconds

SFS Saybolt Furol Seconds

Kinematic Viscosity Centistokes (Centipoises)

Redwood Seconds(Standard)

31 1.00 29 35 2.56 32.1 40 4.30 36.2 50 7.40 44.3 60 10.20 52.3 70 12.95 12.83 60.9 80 13.70 15.35 69.2 90 14.44 17.80 77.6 100 15.24 20.20 85.6 150 19.30 31.80 128 200 23.50 43.10 170 250 28.00 54.30 212 300 32.50 65.40 254 400 41.90 87.60 338 500 51.60 110.00 423 600 61.40 132.00 508 700 71.10 154.00 592 800 81.00 176.00 677 900 91.00 198.00 762 1,000 100.70 220.00 896 1,500 150.00 330.00 1,270 2,000 200.00 440.00 1,690 2,500 250.00 550.00 2,120 3,000 300.00 660.00 2,540 4,000 400.00 880.00 3,380 5,000 500.00 1,100.00 4,230 6,000 600.00 1,320.00 5,080 7,000 700.00 1,540.00 5,920 8,000 800.00 1,760.00 6,770 9,000 900.00 1,980.00 7,620 10,000 1,000.00 2,200.00 8,460 15,000 1,500.00 3,300.00 13,700 20,000 2,000.00 4,400.00 18,400

Kinematic Viscosity (Stokes) =Gravity Specific

(Poises) Visc. Absolute

1 Centistoke = Stoke

100

1 Centipoise = Poise

100

1 Stoke = 100 Centistokes

1 Poise = 100 Centipoises

Centipoises The term “Centipoises” is commonly referred to as a measure of Kinematic Viscosity. Convert centipoises to centistokes by dividing the specific gravity of the solution at the operating temperature.

Plotting Viscosity If viscosity is known at any two temperatures, the viscosity at other temperatures can be obtained by plotting the viscosity against temperature in degrees Fahrenheit on log paper. The points lie in a straight line.

Solids Separators Application Data Solids Separators are used for clarification of liquid flow streams. They are well recognized by industry, having today achieved a “preferred” status. Municipalities and US Governmental Agencies well to recognize tube settlers that are an inefficient version of inclined plate separators. On a worldwide basis, solids settlers have been used for about 20 years in a variety of clarification applications. In short, settlers represent a significant advancement in the state-of-the-art of solids/liquid separation.

Solids/liquid plate separators are now enjoying industry wide acceptance as cost-effective particle removal equipment. They were initially used to remove metal hydroxides from metal finishing wastewaters. his is one of the more difficult solids/liquid separation applications requiring low surface overflow rates coupled with steep angles of plate inclination. From this starting point, plate separators quickly became acceptable equipment in the 1980’s to the power industry, pulp and paper, mining, food processors, chemical processing industries, and municipalities on a variety of particle removal applications. Prospective users of this technology can obtain competitive quotes from several qualified manufacturers who make this equipment marketplace available at attractive prices.

Rather than listing the prospective uses of solids/liquid separators in a comprehensive schedule, it is far easier to indicate the single application that is to be omitted from consideration, and that is secondary clarification of suspended growth biological treatment facilities. In this application, bacteria will (or may) grow on the surface of the separator plates rendering them ineffective. The bacteria will grow so long as dissolved oxygen is present in the waste stream.

For virtually every other clarifier type application, separators may be considered. These include:

1. Primary clarification of sanitary wastewaters, 2. Secondary clarification of fixed growth (rotating

biological surface and trickling filters) biological treatment facilities,

3. Process water softening, clarification, and purification (in competition with and as a substitute for solids contact type clarifiers),

4. Chemical precipitation plants, 5. Mineral processing, 6. Steel and foundry works, 7. Utilities (fossil fuel fired)

A. Fly ash removal systems, B. Coal pile runoff,

8. Metal hydroxide removal, 9. Coal washing and coal slurry separation, 10. Paper fiber reclamation and recovery, 11. Asphalt, sand, and gravel production plants, 12. Cement and lime production plants, 13. Industrial wastewater treatment plants, 14. Hazardous waste treatment facilities, 15. Potable water purification, 16. Cooling water blowdown, and 17. Solids/liquid separation in chemical processing

plants.

Sizing Procedure Solids Separators are sized in the same manner as traditional clarifiers. A surface overflow rate is established at some specific design based upon the settling rates of the particles to be removed. There are three methods that can be used in determining the correct surface overflow rate of a separator. They are:

1. To rely upon published technical data for identical or closely similar applications,

2. By empirical observation and laboratory testing, and/or

3. Theoretical prediction based upon the application of Stoke’s Law.

If the application is of sufficient size and cost, all three methods can be used. In fact, pilot plant testing should be conducted whenever the above three methods fail to produce a consistent surface overflow design rate. Generally speaking, separators provide higher degrees of solids separation than traditional clarifiers at identical surface overflow rates. This permits the utilization of data generated from traditional clarifiers.

Should the solids/liquid separation application require chemical feed and flocculation to enhance separation velocities, then jar tests may be necessary in order to determine both chemical feed dosage and resulting particle separation velocities. Again, pilot plant testing may be desirable depending upon the reliability and availability of design data.

All PuriSep Solids Separators remove rising and settling particles simultaneously. All Separators can be supplied with skimmer mechanisms to remove separated scum. Since the Solids Separators remove rising particles, in addition to the settleable solids, they

may be used where the primary treatment application is removal of scum particles, including oil and/or grease. Therefore, should the treatment application ever change from removing solids to that of removing scum, the same equipment may be used. And if the raw water contains both rising and settling particles, and removal of both types is desirable, the PuriSep Solids Separator is an ideal equipment choice. Furthermore, all Solids Separators may be retrofitted and thereby converted to Dissolved Air Flotation (DAF) treatment should this process modification become desirable. No other separator provides this comprehensive degree of process treatment flexibility, or compatibility.

Table V

Application gpd/sq ft gpm/sq ft Lime Softening

Low Magnesium 2,000 1.39

High Magnesium 1,600 1.11

Alum Coagulation

Turbidity Removal 1,200 0.83

Color Removal 950 0.66

Secondary Clarification

Fixed Growth Systems 1,200 0.83

Suspended Growth*

Contact Stabilization 1,200 0.83

Step Aeration 1,200 0.83

Extended Aeration 1,000 0.69

Separate Nitrification 800 0.56 * Not recommended for inclined plate separators

Sludge Removal Considerations Sludge concentration in the sludge holding compartment will generally range from 1% to 5% suspended solids by weight. This sludge may be thickened to higher solids levels in sludge thickeners.

To be remembered is that the Solids Separator is primarily a particle removal device. Its separated solids or sludge holding capacity is limited. Accordingly, sludge removal is to be accomplished at a rate that matches, or slightly exceeds, sludge separation rates. Accumulated sludge should not be stored in the sludge holding compartment. Most sludge, and virtually all chemical type sludges, will harden on storage. Therefore, sludge removal must match or exceed sludge production.

Solids Separator Selection As a necessary first step prior to separator sizing, the settling velocity of the particles to be removed must be determined. Under the section on Sizing Procedure,

three methods are used to determine the design surface overflow rate.

The first method permits reliance on published data for identical or closely similar applications. For example, the surface loading rates for some common applica-tions are shown in Table V.

Figure 19

The second method requires empirical observation through laboratory testing. The procedure is extremely simple, accurate, and reliable for scale-up design pur-poses. Figure 19 shows a graduate cylinder or a beaker.

1. Prepare apparatus as is indicated above. 2. Introduce thoroughly mixed and fully sus-

pended sample up to the level of Mark 1 and immediately start timer.

3. Stop timer at moment when all suspension in liquid is even or level with Mark 2.

4. Settling velocity is calculated as:

Vp =30H

t= inches/minute

where H = height in inches, and t = time in seconds

5. Repeat steps #2 through #4 two more times and then average the three results

The third method is a theoretical prediction based upon the use of Stoke’s Law. For example, let’s determine the settling velocity of 30 micron sized particles with a

density of 2.1 gm/cc suspended in water @ 20°C. Stoke’s Law is:

Vp =G

18av(dp dc )D2−

where Vp = settling velocity in cm/sec,

G = gravity constant of 980.665 cm/sec2,

η = absolute viscosity of carrier fluid in poises (for water viscosities & densities see Table I, page 15)

dp = density of particle to be removed in gm/cc,

dc = density of carrier fluid in gm/cc,

D = diameter of particle to be removed in centimeters

substituting, we have

Vp =

980.665

18 (0.010050)(2.1 0.998)(3 x10 3 )2− −

= 0.0538 cm/sec, or times 23.628

= 1.271 inches/minute

Having now determined the design rate settling velocity, and knowing the design flow, we can establish a surface overflow rate. For example, let’s say the design settling rate of particles to be removed is 1.271 inch/minute at a design flow of 1000 gpm. From page 1 Equation #1, we know that:

Vp = Q

A, and also that

A = Q

Vp

substituting, we have

A = 1000 GPM

1.271 in / min

= 1262 sq ft required, and

again substituting, we have

A = 1000 GPM

1262 sq ft

= 0.792 gpm/sq ft, or

= 1141 GPD/sq ft overflow rating

On the other hand, if there already exists a separator (or clarifier) with a known separation area and flow, we can determine the settling velocity of the particles that will be removed, disregarding efficiency factors. Let’s say the separator (clarifier) has a separation area of 2,000 sq ft and a flow of 1,000 gpm. Again:

Vp = Q

A

= 1000 GPM

2000 sq ft

= 0.5 gpm/sq ft, or

= 0.6684 feet/minute, or

= 0.8 inches/minute

In other words, an over flow rate of 0.5 gpm/sq ft (720 GPD/sq ft) is the same as and equivalent to the statement that the separator (clarifier) is capable of removing all particles with a minimum settling rate of 0.8 inches/minute. See comparisons in Table VI.

Surface overflow rates, however, should be closely reviewed to prevent non-laminar settling conditions from occurring.

Summarizing, in order to select and rate a separator (clarifier), one must know any two of these three items:

1. Flow

2. Surface Area,

3. Minimum settling velocity of particles to be removed.

In practice, the flow is generally known but the particle settling rates and required surface area are both unknowns. Therein lies the reason for determining settling velocities.

Table VI

Surface Overflow Rate

Settling Rate of Particle Removed

gpm/ sq ft

gpd/ sq ft

m3/ m2 d

inch/ minute m/h

0.25 360 14.7 0.4 0.61

0.50 720 29.4 0.8 1.22

0.75 1,080 44.1 1.2 1.82

1.00 1,440 58.7 1.6 2.43

1.25 1,800 73.4 2.0 3.04

1.50 2,160 88.1 2.4 3.65

As a practical matter, if the particles that require removal have a settling velocity of less than about 0.75 inches/minute, chemical feed and flocculation should be considered. It’s also possible that DAF (Dissolved Air Flotation) may be a cost-effective alternative to a solids separator. Since the PuriSep Separation Systems may include chemical feed, flocculation, and flotation separators, our ability to recommend the most cost-effective treatment system is not encumbered by process or equipment limitations.

Another extremely promising particle/liquid separation process utilizes chemical precipitation technology. This process, called OPCT™ (Optimized

Physical/Chemical Treatment), achieves optimized chemical treatment by careful control of the precipitation process. A detailed description is contained in WSE Publication 995.

Laboratory Testing The WaterSmart Environmental laboratory is available to analyze samples for the purposes of determining settling velocities, conducting “jar tests” for flocculation or emulsion breaking, and selecting chemical feed dosages. Pilot plant testing is also available.

All laboratory testing is provided on a no cost, no obligation basis. Pilot plant testing is available on a

negotiated cost basis depending on complexity, duration, and location of pilot test site.

The following Application Data Sheet identifies both raw water characteristics, and solid/liquid separation requirements. In order to recommend the most cost-effective equipment for the application under consideration, it is extremely important to know the requested data. This is especially true when the flow stream contains a multiplicity of particles, only some of which need to be separated to satisfy removal requirements.

Notes

Application Data Sheet Application: Brief description of solids separator intended use:

Capacities:

A. Design Flow Rate:

B. Peak Flow Rate & Duration:

C. Surface Overflow Rate:

Raw Water Characteristics:

A. Suspended Solids Concentration:

B. Solids Settling Rate:

C. Solids Specific Gravity or Density:

D. Temperature of Liquid: Normal Maximum Minimum

E. pH of Liquid: Normal Maximum Minimum

F. Specific Gravity of Liquid:

Treated Water Requirements:

A. Suspended Solids or Turbidity:

B. BOD or COD:

C. Does NPDES Permit Apply? No Yes Copy Attached No Yes

D. Other:

Continued on the next page

No Cost Laboratory Evaluation: Send (freight prepaid) a two gallon sample of raw feed water along with a small quantity of special treatment chemicals, if any, to our Analytical Laboratory. A wide-mouth, non-breakable, and corrosion resistant (plastic) bottle is preferred. Do not send toxic or otherwise dangerous samples without proper identification of the hazardous waste materials.

Sample (check one)

Has been sent

Will be sent

Will not be sent

Shipping date: via:

Completed Laboratory Report should be sent to:

Company:

Address:

Telephone: Fax:

Attention: Ext:

Email:

WaterSmart Environmental Local Representative:

Notes: 1. The data provided on this Application Data Sheet is our primary source of design

information. The more relevant this information, the faster and more accurate our laboratory work will be. Should we have questions concerning this application, our laboratory staff will contact you.

2. We would like to know a little about the process that generates the solids separator application. A process flow sheet would be ideal. We are not interested in receiving any information you regard as secret or proprietary.

3. Please define as accurately as you can the degree of treatment required to satisfy the application. If EPA or State permit requirements are to be met, please advise the discharge limitations.

4. How would you like to dispose of the separated solids? To what degree does the sludge need to be dewatered?

5. Do you plan to use treatment chemicals? Is pH adjustment a necessity or a consideration? May treatment chemicals be used if they are found to be advantageous to particle separation?

6. Biodegradable samples (food, pulp & paper, etc.) should be preserved before shipment. Please add 1% to 2% by volume of 35% formaldehyde and ship as fast as is possible, preferably an overnight delivery carrier.

Please notify us one day prior to your shipment so we can schedule the immediate testing of your sample on the day received. Do not ship perishable samples on Friday, or preceding a holiday.

Oil/Water Separators

Application Data Oil/Water Separators have been used for the last fifty years to remove floating oil from petroleum processing facilities and their storm drainage systems. The American Petroleum Institute (API) has published oil/water separator design information. The API rec-ommendations have been widely adopted, both in this country and abroad. Oil/water separators that have been designed according to Chapters 5 & 6 of the API Design Manual are called “API Basins” or “API Separa-tors”. They consist of long, shallow, rectangular con-crete built-in-place basins. Their design enables re-moval of all oil particles 150 micron in size and larger. Treated effluent from API basins generally contain about 100 mg/L of “free oil”.

Since 1972, the United States Environmental Protec-tion Agency has significantly reduced the concentration of oil allowable in refinery and other stormwater dis-charges. The API basin cannot comply with the cur-rent treatment requirements. This has led to the de-velopment of more efficient oil/water separation equipment.

For the most part, oil/water separators utilizing corru-gated plates have fulfilled the need for better oil/water separation equipment. Surface oil skimming devices consisting of rotating discs, endless belts, surface skimmers, and similar equipment are not to be con-fused with oil/water separators. Surface oil collection devices remove accumulated surface oil only. They do not remove oil particles entrained in the liquid flow stream. They are, therefore, to be regarded as gross collected oil removal devices. Oil skimming devices are not capable of treating liquid flow to the EPA maximum allowable concentration limitations.

In May, 1979, the API published Bulletin No. 1630, which officially recognized the oil/water separation ca-pability afforded by corrugated plate separators. This publication was followed in 1990 by API Publication 421, which contains design and sizing information on gravity type oil/water separators. In 1992 the EPA published Sand Filter Design for Water Quality Treat-ment of stormwater. In 1994 the US Army Corps of Engineers published an ETL (engineering technical letter) on the selection and design of oil/water separa-tors at Army facilities. Also in 1994 WaterSmart Envi-ronmental issued Publication 194 on Equipment Main-

tenance Wastewater Treatment at Commercial Yards, Garages, and Repair Facilities, Publication 294 on Stormwater Runoff and Washdown Treatment at Automotive Dealership/Service Stations, Publication 394 on A Historical Review of Oil/Water Separator De-signs, and Publication 494 on Stormwater Runoff Treatment at Automotive Wrecking/Salvage Yards. The installation of oil/water separators in this country now numbers in excess of 15,000 to reflect their wide-spread acceptance.

In general, oil/water separators fall into two distinct ap-plications because the raw water characteristics and the treated water requirements both vary. They are:

1. Stormwater with discharge to the environment under an NPDES permit A. Stormwater may contain

a. Free oils/fuels b. Suspended solids w/heavy metals c. Dissolved organics and nutrients

B. Stormwater may be treated using a. Inclined corrugated plate gravity oil/water

separators b. Perimeter trench sand filters

2. Industrial wastewater with discharge to the local treatment plant or zero discharge with water re-cycling A. Industrial wastewater may contain

a. Free oils/fuels b. Chemically emulsified oils due to the use

of detergents c. Suspended solids w/heavy metals d. Dissolved organics and nutrients

B. Industrial wastewater may be treated by us-ing a. High performance oil/water separators

which utilize both filtration and adsorption technologies

b. Chemical precipitation c. Dissolved air flotation

Types of Oily Impurities and Their Separation The oily impurities in stormwater and the wastewaters from refineries and other oil processing operations can be present in four different forms. Each form required a different type of treatment for its effective removal. If more than one form of oil is present, effective removal may require a multiple stage treatment system.

Free Oil

Free oil rises to the surface of the water in which it is contained. The rate of rise of the oil particle is a func-tion of its size and specific gravity as defined by Stoke’s Law. The larger the particle, the faster it rises. Free oil is the most common and usually the predomi-nant oily impurity in wastewater from stormwater and oil processing facilities. Usually more than 95% nu-merically of the oil droplets in a mixture of water and free oil are larger than 50 microns in size. This trans-lates into 99%+ by weight.

Free oil is most economically removed in an oil/water separator utilizing inclined corrugated plates. The oil concentration in the effluent from a PuriSep Oil/Water Separator is less than 10 mg/L provided there are no emulsions and/or dissolved oils in the influent feed.

Mechanically Emulsified Oil

Exposure of a free oil and water mixture to severe tur-bulence will break up the oil into very small droplets, or particles. These particles will range in size from 10 to 40 microns and are called a mechanical emulsion of oil in water. In the transfer of oily wastewater, centrifugal pumps, flow restricting orifices, short radius elbows, and other appurtenances generating turbulence should be avoided, where possible, to avoid the formation of mechanically emulsified oil.

Mechanically emulsified oil may be economically re-moved in an oil/water separator utilizing Dissolved Air Flotation (DAF) to assist particle separation. In this process, micron sized air particles attach to the me-chanically emulsified oil particles and float them to the liquid surface. Since these oil particles are so small, they do not possess a sufficient rise rate, or separation rate, to permit simple gravity separation. Mechanically emulsified oils may also be removed by oleophilic me-dia filtration/adsorption.

Should free oil be present in concentrations greater than 2,000 mg/L along with the mechanically emulsi-fied oil, removal of the free oil will be necessary before DAF removal of the mechanically emulsified oil is at-tempted. The presence of large quantities of free oil retards the ability of the DAF process to remove the mechanically emulsified oil particles.

The oil concentration in the effluent from this system is less than 10 mg/L provided there exists no chemically emulsified and/or dissolved oil in the influent, and pro-vided further that the free oil concentration is less than the 2,000 mg/L.

Chemically Emulsified Oil Chemically emulsified oil particles are less than 1 mi-cron in size. They do not rise to the surface of the wa-ter in which they are contained, no matter how much rising time is allowed. Chemically emulsified oil in wastewaters is usually due to the presence of deter-gents or alkali contaminants.

Chemically emulsified oil may be economically re-moved in a two-stage PuriSep separation system consisting of a first stage CorruFloc flocculator fol-lowed by a second stage DAF flotation separator. The purpose of the flocculator is to “break” the chemical emulsion so that the small micron sized particles can agglomerate together forming a floc of sufficient size to permit air attachment and subsequent removal by flota-tion. Should free oil be present in concentrations greater than 2,000 mg/L along with chemically emulsi-fied oil, removal of the free oil will be necessary before DAF removal of the chemically emulsified oil is at-tempted. The presence of large quantities of free oil retards the ability of the DAF process to remove the chemically emulsified oil. While it is theoretically pos-sible to have both types of emulsified oil, that is, both mechanically emulsified and chemically emulsified oils, present in the same waste stream, as a practical mat-ter, only one type will usually be present. Should both be present, or should the raw water characteristics change from one type of emulsion to the other, the chemically emulsified oil treatment system will also remove the mechanically emulsified oil particles. The oil concentration in the effluent from a DAF system is less than 10 mg/L provided there is no dissolved oil in the influent, and provided further that the free oil con-centration is less than 2,000 mg/L. Chemically emulsi-fied oils may also be removed by adsorption or chemi-cal precipitation technologies. Granular activated car-bon (GAC) is a commonly used adsorber to remove these oils. Likewise, if powdered activated carbon (PAC) is used with chemical precipitation, successful emulsified oil removal can be accomplished.

The choice of technologies to be used in the removal of chemically emulsified oils depends on:

1. Raw water characteristics 2. Treated water requirements 3. Size of application

Since several technologies can be considered, the most cost-effective selection will differ depending on the above variables. Therefore it is quite appropriate to cost out each of the several alternatives available.

Dissolved Oil

Dissolved oil.is no longer present in the form of dis-crete particles. Therefore, it cannot be removed from the flow stream with an oil/water separator. Short-chain hydrocarbons such as hexane have limited solu-bility in water, but long-chain hydrocarbons are practi-cally insoluble. Aromatic compounds such as benzene

and toluene, phenol, etc., are quite soluble in water and are considered as oil in most analyses.

Should dissolved oil be present in quantities sufficient to warrant removal, the use of air stripping, activated carbon absorption columns, biological treatment, or all three may be considered. Selection of the most cost-effective dissolved oil removal system depends upon the kinds of oils present, their respective concentra-tions, and required removals.

Treatment Options A classification of oil removal capabilities is given in Table VII below:

Table VII

Type of Oily Contaminants

Type of Separation System

Free Oil Oil/Water Separator Mechanically Emulsified Oil

PuriSep DAF Separator or Con-tamAway I Adsorption Separator

Chemically Emulsified Oil

Combination CorruFloc Flocculator & PuriSep DAF Separator or Con-tamAway II Adsorption Separator

Mixture of Above

Multi-Stage Flocculator and/or Separators

Emulsion Breaking The breaking of a mechanical emulsion is generally not required since the oil particle is sufficiently large to permit air attachment and subsequent removal by flota-tion.

The breaking of chemical emulsions is, on the other hand, quite necessary because of the extremely small size of the emulsified oil particle. Unfortunately, the “breaking” of a chemical emulsion is more art than sci-ence. Oxidizing agents such as ozone and potassium permanganate have been employed. For the most part, chemical emulsions are most quickly and eco-nomically broken by the use of alum along with various polyelectrolytes. In determining the most effective method to treat each oily waste, jar tests are usually performed. Jar tests establish the chemicals and/or coagulant aids required, the chemical feed dosages, and the degree of treatment achievable.

Often coagulation with alum alone will break simple emulsions. The aluminum ion neutralizes any anionic surface charges on the suspended particles or oil drop-lets. The excess aluminum ion forms hydrous oxide, which provides a surface on which the oil is absorbed. Floc formation is the result of alum’s reaction with the alkalinity normally present in most wastewaters.

The dosage of alum required to achieve emulsion breaking and floc formation will vary from one type of wastewater to the next, but generally will be in the range from 90 to 180 mg/L (0.75 lbs to 1.5 lbs alum per

1,000 gallons wastewater). Alum solution feed strength is usually between 5 and 10%.

The use of polyelectrolytes usually enhances floc for-mation and shear resistant qualities, both of which are necessary for optimum flotation separation. The dos-age of polyelectrolytes is generally in the 1 to 2 mg/L range at a concentration between 0.1 to 0.2%. When polyelectrolytes are used along with alum, the alum is added upstream from the polyelectrolyte feed location to permit alum-water reactions to take place. Difficult-to-break emulsions frequently respond to calcium chlo-ride treatment prior to flocculation. Occasionally, pH adjustment may be necessary prior to chemical addi-tion.

A chemical analysis of an oily wastewater sample does not by itself indicate the quantity of alum or supplemental chemicals needed for successful treatment. The coagulant demand is largely dependent upon certain chemical and physical properties. These properties include the quantity and types of particles in suspension and their zeta potential, pH conditions, alkalinity, emulsifying agents, bactericides, and other materials that may be present. Jar tests are the only reliable method of determining chemical feed requirements on a consistent basis and are therefore recommended on any waste stream that contains chemical emulsified oil that requires removal.

Sizing Procedure Oil/Water Separators are sized in the same manner as traditional clarifiers. A surface overflow rate is estab-lished at some specific design based upon the rising rates of the oil particles to be removed. There are three methods that can be used in the determination of the appropriate surface overflow rate of a separator. They are:

1. To rely upon published technical data for identi-cal or closely similar applications,

2. By empirical observation and laboratory testing, and/or

3. Theoretical prediction based upon the applica-tion of Stoke’s Law.

If the application is of sufficient size and cost, all three methods can be used. In fact, pilot plant testing should be conducted whenever the above three methods fail to produce a reliable overflow design rate. Generally speaking, separators provide higher degrees of oil/water separation than traditional clarifiers at identi-cal surface overflow rates. This permits the utilization of data generated from traditional clarifiers.

Should the oil/water separation application require chemical feed and flocculation to enhance separation velocities, then “jar tests” may be necessary in order to determine both chemical feed dosage and resulting particle separation velocities. Again, pilot plant testing may be desirable depending upon the reliability and availability of textbook design data.

Sources of Oily Waste Table VIII

SIC Industry Description

20 Food Processing Natural fats and oils resulting from animal and plant processing, including slaughtering, cleaning, and by-product processing.

27 Textile Manufacturing Oil and grease from wool scouring and cotton finishing operations.

29 Petroleum Light and heavy oils resulting from the production, refining, storage, trans-portation, and retailing of petroleum and petroleum products.

33 Primary Metals Rolling and drawing oils resulting from mill rinses and scale pit effluents of ferrous and non-ferrous metals manufacture.

34, 35, 37,

Fabricated Metals, Machin-ery, & Transportation Equip-ment

Grinding, lubricating, stamping, and cutting oils employed in ferrous and non-ferrous metal fabricating and rinsed from parts in clean-up operations.

-- Cooling & Heating Dilute oil-containing cooling water, from leakage in pumps, condensers, heat exchangers, etc.

-- Transportation Services Oils and greases generated during the repair, maintenance, fueling, and cleaning of transportation vehicles and facilities.

True Inclined Plate Separators Several marketplace products are represented as in-clined plate separators. Most of these are not, how-ever, true inclined plate separators.

An example is shown in Figure 20. This separator uses vertical plates with inclined right angle ribs. Note that a once separated particle must pass through the active

flow stream several times (particle reentrainment) en route to its destination. Also note that rising and set-tling particles are migrating to their respective destina-tions on the same side of the plate.

A kissing cousin of the Figure 20 separator utilizes ver-tical plates with inclined corrugations rather than in-clined ribs. See Figure 21.

Figure 20

ISOMETRIC VIEW (Single Plate)

ISOMETRIC VIEW (Slant Rib Plate Pack)

View A-A 90°

Common hydraulic passageway between adjacent plates

Common particle migration zone

Potentially reentrained rising particles

¾" ±

1½ " ±

Potentially reentrained settling particles

Flow path

Figure 21

About the only positive with this design is a lower Reynolds Number than the inclined rib design in Figure 20. The same deficiencies regard-ing particle migration, however, still exist in the above design. Another misrepresented inclined plate separator is shown in Figure 22.

Figure 22

ISOMETRIC VIEW (Single Plate)

View A-A

Common hydraulic passageway between adjacent plates

Common particle migration zone

Potentially reentrained settling particles

Flow path

Potentially reentrained rising particles

ISOMETRIC VIEW (Slant Rib Plate Pack)

¾" ±

Even though the plates are indeed inclined, their lengths have been arbitrarily shortened to accommo-date ease of assembly. Note that once separated par-ticles (those having reached a boundary layer) must again pass through the active flow stream (particle re-entrainment) en route to their liquid surface or vessel floor destinations. This particular closed-in design util-izes polypropylene (oleophilic) plastic. The manufac-turer claims both “self-flushing” and “easy-to-clean” characteristics and therefore minimum maintenance. These overt claims are scoffed at by end users of the product.

Figure 23

Yet another misrepresentation is the characterization of the following type separator as an inclined corru-gated plate separator. See Figure 23. This product uses inclined plates that have corrugations. The plates are inclined but the individual corrugations are horizon-tal. As the oil rises to the ceiling of each plate it must thereafter slide up additional corrugations to reach the surface of the vessel. The solids that separate out will tend to fill each of the corrugations as shown in Figure 24.

Figure 24 After solids have filled the corrugations, the flow must become turbulent as the flow passageway at cross section A is much greater than the passageway at sec-tion B. In order to maintain laminar flow, the flow pas-sageway must remain uniform in cross section. The oil removal capacity of this design mostly disappears as soon as sludge fills the corrugations. In a true corru-gated plate separator, the plates and the corrugations are both inclined thereby permitting the sludge to slide

to the bottom thus maintaining a uniform cross section. As a practical matter, if a separator design accom-plishes effective separation and removal of settling par-ticles, it should also achieve excellent removal of rising particles. Both features ought to be considered in the selection of a preferred design.

A true inclined plate separator uses full-length plates as shown in Figure 25.

Figure 25 Full-length plates absolutely eliminate the possibility of particle reentrainment since once-separated particles stay separated as they migrate along a boundary layer (immediately adjacent to plate) and therefore cannot become reentrained thus maximizing effluent quality.

Figure 26 exhibits the removal of rising and settling particles on opposite sides of plates.

Figure 26

Note: If a rising oil particle touches a settling sludge particle, the resulting oily sludge particle may well ex-hibit “neutral gravity” and thus resist effective separa-tion to the detriment of effluent quality. Achieving sepa-ration of rising and settling particles on opposite sides of each inclined plate greatly enhances separator per-formance when both types of particles are present as they usually are.

True inclined plate separators use a non-oleophilic ma-terial, like fiberglass, with smooth surfaces on each side to promote particles to slide to their destinations at a generous angle of repose (inclination angle). The less the angle of repose the less expensive the separa-tor. When in doubt the use of a 55° minimum angle of

inclination and fiberglass plates will insure excellent particle migration along the plate surfaces at Reynolds Numbers up to 500.

Figure 27

By comparison, the use of inclined plates at less than 45° or the use of oleophilic plates at all angles of incli-nation represents designs that are susceptible to plug-ging. Plate spacing at ¾" will virtually eliminate the plugging whereas plate spacing at ½" or even ¼" will accelerate this condition and the associated loss of effluent quality.

Figure 28

True inclined plate separators may use either flat or corrugated plates. Flat plates accomplish singular par-ticle coalescence in the direction of particle migration along the plate surface as a faster migrating particle catches up to a slower migrating particle. Corrugated plates accomplish dual particle coalescence. In addi-tion to the particle coalescence achieved in the direc-tion of particle migration (identical to flat plates), corru-gated plates accomplish coalescence in the direction of fluid flow. As a particle settles (or rises) to a boundary layer, it thereafter slides to a trough as it migrates to its destination. It’s the sliding into a trough that constitutes the additional coalescence than that exhibited by flat plates.

True inclined plate separators may be either parallel flow or cross flow. In a parallel flow separator, the di-rection of fluid flow is cocurrent or countercurrent to the direction of oil particle migration as shown in Figures 27 and 28, respectively. In a cross flow separator, the fluid flow is perpendicular to the direction of particle migration as shown in Figure 29.

Figure 29 In a cocurrent or countercurrent separator, some parti-cle reentrainment can occur as a result of particles passing through an active flow stream as is shown in Figures 30 and 31.

Figure 30

Particle reentrainment always degrades separator per-formance. This phenomenon cannot occur in a cross flow type separator provided the Reynolds Number is less than 500 and full-length plates are used.

True inclined plate separators use gravity separation rather than adsorption to remove oil. Stoke’s Law es-tablishes the separation rate of discrete particles rela-tive to size of particles, specific gravity differential be-tween the particles and their carrier fluid, and the tem-perature of the carrier fluid. Several manufacturers of oil/water separators include devices such as polypro-pylene (oleophilic) perforated tubes, polypropylene filament filters, and even polypropylene plates all of which accomplish removal of oil through the mecha-nism of adsorption rather than gravity separation. Notwithstanding this fact, adsorption type separators

are routinely marketed under the name of “enhanced gravity”, “petro-pak coalescer”, or “coalescer” so as to disguise the true process of oil removal.

Figure 31

Some 15 companies sell polypropylene based (oleo-philic) adsorption type products to industries and remediation contractors to soak up (through the phe-nomenon of surface adsorption) oily spills, drips, and/or contaminants. The mechanism of oil removal is identical whether the oleophilic product is housed in an oil/water separator or otherwise. To call an adsorption device by a name that implies gravity separation is a clear misrepresentation of the technology being used. As a backdrop, oleophilic adsorption devices are ex-tremely inexpensive to make and most effective in re-moving oily particles but their useful life is extremely short. These devices must be replaced when their ad-sorptive capacity is reached, or when they otherwise become plugged with solids. Gravity type separators are more expensive initially but they continue to per-form year after year with only minimal maintenance required. Short term cost savings are invariably out-weighed by long term maintenance expenses. There-fore, if the oil/water separator design under considera-tion utilizes oleophilic polypropylene, the end user should be prepared to accomplish heavy maintenance to maintain effluent quality.

Plate Inclination Angle Projected plate separation area is determined by multi-plying the actual plate area by the cosine of the angle of inclination. As examples, a 4 ft x 8 ft plate inclined at 60° has a projected separation area of 16 sq ft as shown in Figure 32. Similarly, the same 32 sq ft plate has greater area projections at lesser angles of inclina-tion as is shown in Table IX.

Figure 32

Table IX

Projected Separation Area Angle of Inclination of 32 sq ft Plate

Separation Area in sq ft

60° 16.00

55° 18.35

45° 22.63

30° 27.71

The lesser the angle of inclination the more suscep-tible the separator becomes to plugging with solids. If the angle of inclination is insufficient to overcome the angle of repose of the particles, the separator will foul after but a short operating period. Therefore, even though lesser angles possess greater separation ar-eas, the lesser angles also result in heavy mainte-nance cleaning requirements. Even though a separator with greater angles of inclination posses lesser pro-jected plate areas, the associated one time increase in cost far outweighs persistent increased maintenance costs associated with lesser angles of inclination. An inclination angle of 55° always exceeds the angle of

repose and is therefore recommended as a minimum angle.

Free Oil Removal In order to design any oil/water separator four things are required:

1. Design temperature of operation. As the tem-perature drops, liquid viscosities increase thereby decreasing particle separation rates. Performance of oil/water separators therefore decreases with falling liquid temperatures.

2. Specific gravity differential (or density differ-ence) between the oil to be removed and the carrier liquid.

3. The diameter of the minimum sized oil particle to be removed.

4. The design flow rate at the design temperature of operation.

A typical example could be as follows: 1. Design temperature = 20° C 2. Gravity differential = 0.10 gm/cc 3. Diameter of minimum sized oil particle to be

removed = 60 microns 4. Flow is 575 GPM

Step #1 requires us to calculate the separation velocity of the minimum sized oil particle to be removed. Stoke’s Law can provide this:

Vp =G

18av(dc dp )D2−

where Vp = separation velocity (rise rate) of oil parti-cle to be removed.

G = gravity constant of 980.665 cm/sec2,

η = absolute viscosity of carrier fluid in poises (for water viscosities & densities see Table I)

dp = density of particle to be removed in gm/cc,

dc = density of carrier fluid in gm/cc D = diameter of particle to be removed in

centimeters. substituting, we have

Vp = 980.665

18 (0.010050)(0.10) (6 x 10 3 )2−

= 0.0195 cm/sec, or times 23.628 = 0.461 inches/minute

Step #2 requires us to calculate the effective surface area of the separator to remove those oil particles with a separation velocity of 0.461 in/min. From page 1 Equation #1, we know that:

Vp = AQ , and also that

A = Q

Vp

substituting, we have

A = 575 gal / min

0.461 in / minx

12 in

1 ftx

1 cu ft

7.481 gal

= 2,000 sq ft of effective separation area required to remove 60-micron oil particles at the design conditions

Now in order to predict the concentration of oil in the effluent of the preceding oil/water separator, two things are required: initial oil concentration in water, and oil particle distribution at the design conditions. For the purpose of an example, let’s assume the fol-lowing:

1. Oil concentration = 1,000 mg/L, and 2. Oil particle distribution at the design conditions

as follows: (typical stormwater runoff distribu-tion)

Table X

Oil Particle Size in Microns

% Concentration by Weight

greater than 150 90 150 to 120 5 120 to 90 2.5 90 to 60 2 less than 60 0.5

Since the example oil/water separator will remove all oil particles 60 micron in size and larger, 99.5% of the influent oil will be removed. 99.5% removal of 1,000 mg/L initial oil concentration leaves 5 mg/L oil in the effluent.

There are no short cuts to these projected plate area and treatment efficiency calculations. Therefore, a supplier cannot technically guarantee the effluent qual-ity of an oil/water separator unless the following are either known or presumed:

1. Water temperature 2. Oil and water density difference 3. Minimum oil particle removal size 4. Design water flow rate 5. Inlet oil concentration 6. Oil particle concentration (by weight) distribu-

tion characteristics.

The Relationship between Projected and Effective Plate Separation Area The Stoke’s Law calculations to determine projected plate separation area do not take into account the effi-ciency of a separation device. Flow bypassing, flow distribution across the inlet face, treated water collec-

tion, and laminar flow all impact on particle removal efficiency. A well designed unit such as a PuriSep oil/water separator will exhibit 95% particle removal efficiency whereas a poorly designed unit may exhibit but 35% efficiency. Reputable oil/water separator manufacturers will have compared the results of their separators against the theoretical results to arrive at the efficiency of operation. This is referred to as an efficiency factor.

The efficiency factor multiplied by the projected plate separation area equals the effective separa-tion area.

Some manufacturers use a term called a safety factor to reflect the presence of an apparent excess amount of separation area. A 2.86 safety factor, for example, is the same as an efficiency factor of 35% since 100 ÷ 35 equals 2.86.

Oil/water separators that use corrugated fiberglass (non-oleophilic) plates at a high angle of inclination (55° minimum) will exhibit excellent efficiencies for ex-tended periods (many years). Oil/water separators which use flat plates, oleophilic plates (polyethylene, polypropylene, or Teflon), rough surface plates (PVC) or low angles of inclination (less than 55°) will exhibit lesser efficiencies over a much shorter time frame measured in months rather than years. Lesser efficien-cies always require more expensive maintenance pro-cedures over the life of the installation. Therefore greater efficiencies are extremely important to achieve and may be considered as a specified feature.

Surface Area This term traditionally meant the square feet of floor or separation area possessed by a clarifier. As inclined plate separator technology became introduced, this term was properly applied to the sum of the projected areas of each plate.

As competition increased in the marketplace, oil/water separator vendors quickly learned that they could suc-cessfully claim credit for the much greater actual sur-face area of the plates, polypropylene tubes, and poly-propylene mesh filters as a substitution for the much smaller projected area that, in fact, existed. This con-tinuing ruse is highly successful for two reasons:

1. The sales and marketing literature of the vendor invariably pays homage to conformance with API design literature, Stoke’s Law, laminar flow, and to the concept of greatly increased separa-tion area possessed by inclined plates. The above references refer to classical particle separation technology. The reader therefore rightfully expects the represented technology to be delivered in the equipment.

2. The actual product, however, is fraudulently supplied on the basis that the square feet of separation area is satisfied by the actual sur-

face area of the plates, tubes, or filters rather than the projected area. One company even claims both sides of each plate in arriving at projected area, a scientific impossibility. Since the end user, or their consulting engineer, usu-ally does not detect this vital substitution, wide-spread abuse of the term surface area has oc-curred. In an effort to mask the real meaning, most vendors frequently use the term coalesc-ing area to imply projected separation area while secretly meaning actual surface area. The obvious motive for such practice is to supply less equipment (less projected plate area) for the price quoted.

Square Feet of Projected Area Per Cubic Foot of Media Rather than disclose how square feet of projected separation area is arrived at, several vendors simply claim a specific projected area without supporting cal-culations. To assist the designer in determining the correct projected area, the following figure shows a cubic foot of inclined plate media with ¾ inch plate spacing and a 55° angle of inclination.

Figure 33 The above figure shows 17 plates, 16 of which project area. The actual plate area multiplied by the cosine of the angle of inclination equals the projected area or:

16 x cos 55° = 9.17 sq ft/cu ft

The following table was prepared showing projected area per cubic foot of media at different plate spacing and angles of inclination as follows:

Table XI

Sq Ft/Cu Ft at Plate Spacing of Plate Angle of Inclination ¼" ½" 3/4" 2" 40° 36.77 18.38 12.25 4.59

45° 39.94 16.97 11.31 4.24

50° 30.85 15.42 10.28 3.85

55° 27.53 13.76 9.17 3.44

60° 24.00 12.00 8.00 3.00

Plate thickness has not been taken into consideration in arriving at the above figures. Plate thickness would decrease the above plate areas by about 5%. Those vendors who characterize separation areas as sq ft/cu ft of media routinely misrepresent actual projected ar-eas by factors of 2 to 6 times. In order to approximate the actual projected plate area one must know the:

1. Cubic feet of media 2. Plate angle of inclination 3. Plate spacing (vertical distance between adja-

cent plates)

The Relationship between Capacity and Performance If the page example separator were required to treat a two year storm event of 1294 gpm, or 2.25 times as much flow as the example, what sized oil particles would be removed and what would the oil concentra-tion be in the effluent?

Solution: At the new flow rate, the surface overflow rate becomes 1294 ÷ 2,000 sq ft or 0.647 gpm/sq ft This is the same as saying the separator will remove all particles with a separation velocity of:

Vp = Q

A

= 0.647 gpm / sq ft

1 sq ftx

1 cu ft

7.481 galx

12 in

1 ft

= 1.038 inches/minute, or divided by 23.628

= 0.044 cm/sec

Substituting this particle separation velocity in the pre-vious Stoke’s Law calculation and solving for the oil particle diameter results in a 90-micron oil particle re-moval size. This means all particles 90 microns and larger will be removed or 97.5% of the influent oil con-centration (refer again to the oil concentration particle distribution). The remaining 2.5% amounts to an efflu-ent oil concentration of 25 mg/L (2.5% of 1000 mg/L).

Knowing the relationship, then, between capacity and performance enables the designer to specify a cost-effective oil/water separator. For example, in a refinery

rainwater runoff application, the water flow and oil con-centration may vary over a considerable range. The initial flow might be 575 gpm. The initial “first flush” oil concentration is generally high, usually about 2,000 mg/L, representing oil spilled on operating platforms and pads. A typical initial oil particle distribution is:

Oil Particle Size in Microns

% Concentration by Weight

greater than 150 90 150 to 120 5 120 to 90 2.5 90 to 60 2 less than 60 0.5

At this flow (at the example design conditions of tem-perature and density differential), the separator would remove all particles 60 micron in size and larger or 99.5% of the inlet oil concentration resulting in 0.5% of 2,000 mg/L or 10 mg/L oil in the effluent.

After a few minutes, the rainfall rate typically increases significantly, say 2.25 times to 1294 gpm, as the oil concentration and oil particle size simultaneously de-crease. A new oil concentration might be 100 mg/L (reflecting that the majority of the spilled oil has already been rinsed away) with a particle distribution of:

Oil Particle Size in Microns

% Concentration by Weight

greater than 150 50 150 to 120 25 120 to 90 15 90 to 60 7.5 less than 60 2.5

At this flow, as we have seen, the separator will re-move all particles 90 microns in size and larger or 90% of the inlet oil concentration resulting in 10% of 100 mg/L or again 10 mg/L oil in the effluent.

In this last example, the flows and oil concentrations varied considerably, but the effluent oil concentration remained essentially the same. Knowing the relation-ship between capacity and performance can prevent the design of excess capacity under certain applica-tions. Designing for a six-month or two-year storm condition is appropriate. Designing for a 25-year event represents excess capacity.

Oil Particle Size to be removed The marketplace sales literature on oil/water separa-tors contains differing claims on oil particle size re-moval capabilities. Some recommend 60 micron, oth-ers 20 micron, and some claim 10 micron. The subject is important because the rate of separation of a particle is directly proportional to the square of its diameter. A 20 micron sized particle, for example, would require

nine times as much projected plate separation area to remove as a 60 micron sized particle because 602 ÷ 202 = 9. Nine times the separation area translates into about 7 times the cost of such a separator.

The essential question to be addressed is what size particle requires removal in order to achieve the quality of effluent desired. In this regard, the wastewaters to be treated generally fall into the two categories of stormwater and industrial wastewater.

Stormwater can be treated to less than 10 mg/L oil in effluent by removing oily particles down to 60 micron in size since particle size distribution curves reveal the existence of 99.5% of oil particles equal to or larger than 60 microns. The US Army Corps of Engineers, US Navy, and US Air Force all target 60 microns as the removal requirement. Some state agencies (i.e., De-partment of Ecology, State of Washington) also identify 60 microns as the size to be removed.

Industrial wastewaters generally contain chemically and/or mechanically emulsified oils due to the pres-ence of detergents and pumping, respectively. Emulsi-fied oils cannot be effectively removed in a gravity type oil/water separator because oily particle sizes are very small, on the order of 0.1 to 5 microns. Many, if not most, of these particles are so small that Brownian movement absolutely prevents their removal through gravity separation. Gravity type separators are there-fore ineffective in treating industrial wastewaters gen-erated from floor cleaning, car/truck washwater, equipment maintenance, and similar activities. In order to remove the oily contaminants from these wastewa-ters, the use of adsorption or chemical precipitation technologies is necessary. WSE Publication Nos. 894 and 994 address adsorption treatment and WSE Publi-cation Nos. 195, 1194, and 1294 address chemical precipitation treatment.

Without exception, every vendor claiming removal of oily particles to 20 micron in size and smaller uses oleophilic (oil-loving) polypropylene plastics. While it is true that the mechanism of oily particle removal is ad-sorption rather than gravity separation, this type of separator is still ineffective because of the extremely small size of a majority of particles (less than 5 micron in size). Therefore, if the oil/water separator effluent must achieve high levels of performance, neither grav-ity nor oleophilic adsorption type separators ought to be considered.

American Petroleum Institute’s Susceptibility to Separation Test (S.T.S. Number) This test is an actual measurement of the separation rate possessed by rising particles. It permits the accu-rate determination of the required surface overflow rate of a separator to achieve a specific degree of oily wastewater treatment. It must be performed at the very site where the samples are taken. The time re-

quired to transport the samples to a laboratory will change the composition of the sample, and thus the results.

Figure 34 1. Prepare apparatus as shown in Figure 34. Be

sure the separator funnel is clean. Just before performing the S.T.S. test, rinse the funnel with detergent water to wet the interior surfaces. This prevents oil from clinging to the funnel’s wall.

2. Introduce thoroughly mixed (uniform) oily wastewater sample and fill separator funnel to top.

3. Wait two (2) minutes to disseminate initial tur-bulence (time with stopwatch). While oil parti-cle separation is taking place, the funnel should be protected from temperature disturbances. Position the funnel in an area where there ex-ists no wind, or nearby sources of heat. If the oily wastewater being sampled is at a tempera-ture significantly different (say 10°C) from am-bient air, a towel or other type of insulation should be wrapped around the funnel.

4. Draw off and discard 50 cc of sample to waste the settleable solids. Reset timer to 0.

5. Wait five (5) minutes.

6. Draw 600 cc of sample into a clean sample container and send to the laboratory for oil analysis by solvent extraction, infrared spec-troscopy, or other API (American Petroleum In-stitute) test method.

7. Oil particle separation velocity is calculated as follows:

Vp = H

t

where H = height in inches of 600 cc sample, and t = time in minutes If, for example, H = 6 inches, then

Vp = 6

5

= 1.2 inches/minute, which converts to = 0.75 gpm/sq ft surface overflow rate, or = 1,080 GPD/sq ft surface overflow rate

Should the foregoing laboratory test result show an oil concentration of 40 mg/L, this is the projected per-formance level of a separator operating at the surface overflow rate indicated.

8. Repeat steps #1 through #7 except wait for ten (10) minutes. Determine projected perform-ance at this surface overflow rate (1/2 the sur-face overflow rate of the prior example).

Repeat as necessary at additional wait times in order to establish the desired design surface overflow rate.

Treatment of Mechanically Emulsified Oil This type of oily contaminant is relatively easy to re-move in a DAF separator. If the emulsified oil concen-tration is less than 2,000 mg/L, it may be removed at a surface overflow rate of 0.5 gpm/sq ft (corresponds to 720 GPD/sq ft). Air bubbles attach to the emulsified oil particles imparting an apparent density of about 0.8 gm/cc. The resulting air-oil particle exhibits a rapid rate of rise. The limiting factor controlling liquid flow in maintaining laminar flow conditions. At a surface over-flow rate of 0.5 gpm/sq ft, the PuriSep DAF separator is operating at its maximum hydraulic rate.

If free oil is present at a concentration of 0.2% or more, or if suspended solids are present at a concentration of 0.1% or more, removal of these other particles in a first stage gravity separator is indicated.

Mechanically emulsified oil may also be removed by oleophilic adsorption technology.

Treatment of Chemically Emulsified Oil These emulsions consist of oil particles less than 1 micron in size. In order to remove these tiny particles,

the emulsion must be broken through chemical addi-tion. pH adjustment might also be evaluated for emul-sion breaking. After the emulsion is broken, the parti-cles are flocculated and thereby agglomerated to a size large enough for air particle attachment. The treatment, then, consists of a first stage DAF separa-tor. Chemical feed is also provided.

The determination of the proper chemicals to use is accomplished through jar tests. Additional laboratory testing is required to determine the presence of free oil and suspended solids. As before, the maximum sur-face overflow rate of the CorruFloc flocculator/DAF separator is 0.5 gpm/sq ft. If either 0.2% free oil and/or 0.1% suspended solids are present, a first stage grav-ity separator is required.

Chemically emulsified oils may also be removed through adsorption media such as granular activated carbon.

Treatment of Dissolved Oil Should dissolved oil be present in the oily wastewater in sufficient quantities to warrant removal, the choice of equipment is between granular activated carbon ab-sorption contactors, biological treatment, or air strip-ping. The most cost-effective removal method de-pends on the particular dissolved oil(s) present, the concentration of the dissolved oils, & the size of the treatment application. Generally, the lower the concentration of dissolved oil, the less biodegradable the oil, and the lower the flow rate, the more activated carbon treatment is indicated. Conversely, the higher the concentration, the more biodegradable the oil, and the higher the flow rate, the more biological treatment is indicated.

In some applications, first stage biological oxidation followed by second stage activated carbon absorption is the most cost-effective solution. In all instances, laboratory testing is necessary prior to the design of a dissolved oil removal system.

Retrofitting - Upgrading Existing Inadequate Oil/Water Separators Since inclined plate modules can be built to accommo-date almost any vessel, existing inadequate oil/water separators and clarifiers can usually be converted into high performance units. The vessel itself must be structurally sound as a starting point. Then adequate access to the vessel must be provided if not already present.

Figure 35 Figure 35 shows a somewhat typical baffled clarifier. It can be upgraded by:

1. Removing existing baffles 2. Adding inlet and outlet piping 3. Adding sludge collection and withdrawal system 4. Adding oil withdrawal system 5. Adding inclined plate module

The retrofitted clarifier is shown in Figure 36.

Figure 36 Many thousands of phony gravity type oil/water sepa-rators have been sold of the type shown in Figure 37.

These units actually accomplish most of their oil re-moval through polypropylene mesh adsorption tech-nology rather than gravity separation notwithstanding

sales literature copious endorsements of the principles of Stoke’s Law, laminar flow, and low Reynolds Num-bers. Once the oleophilic filter becomes saturated with oil and/or plugged with solids, it is somewhat common practice to remove this element without replacing it. Subsequent performance is then inadequate because of a gross insufficiency of plate separation area. For good measure, the plates may be inclined but the cor-rugations are horizontal resulting in a module partially plugged with solids.

Figure 37 Inserting module components through the access manway may retrofit this type of separator. See Figure 38.

Figure 38 Adding a second module as shown in Figure 39 can actually further modify this type of separator.

Figure 39

Coalescers Many oil/water separator manufacturers frequently use the term coalescer in the description of their product. This is but a popular buzzword in the marketplace as it is invariably used incorrectly. Various manufacturers use oleophilic polypropylene mesh material and tubes to adsorb oily particles. Still others use polypropylene plates to adsorb oily particles. Not to be outdone, a few manufacturers even employ non-oleophilic plastics (i.e., PVC) while ironically claiming the misguided benefits associated with the use of this material. In every such instance the manufacturer incorrectly uses the term “coalescer” in describing its product because of the extremely favorable impression this word has taken on in the representation of particle/liquid separa-tion. If the truth be known, every time a manufacturer uses this term it should be perceived as a red flag and a caution against the use of the coalescer product rep-resented.

Sludge Removal Considerations Sludge concentration in the sludge holding compart-ment will generally range from 1% to 5% settled solids by weight. It is to be remembered that the PuriSep Oil/Water Separator is primarily an oil-removing device. Its sludge holding capacity is limited. Accordingly, a sludge removal requirement is to waste or remove sludge at a rate that matches, or slightly exceeds, the rate of sludge accumulation. Accumulated sludge may not be stored in the sludge holding compartment. Most sludges, and virtually every chemical type sludge, will harden on storage. Many oil/water separators have become inoperative because sludge removal was not routinely practiced.

Maintenance The principal feature of the stormwater separator is to separate and temporarily store various petroleum hydro-carbons, oil/fuels, and lubricants that may be contained

in the water to be treated. A secondary feature of the separator is to separate and temporarily store sand, grit and other settleable materials that may be washed into the drainage system. The fuels and oils separated in the separator will form a layer on the top surface. An oil withdrawal pipe may be used to remove the separated oil on a periodic basis. If supplied with your unit, an adjustable floating oil skimmer automatically removes floating oil from the separator compartment and discharges it into an adjacent waste oil storage tank or compartment. In the event of a known spill the separator and the waste oil collection tank, if provided, should be checked immediately to determine their unused capacity. The oil may be removed by contracting with a licensed disposal service company. The settled solids should be removed periodically. The sludge can be removed in the same manner as the oil, namely by hiring a disposal service company. The disposal of sludge is regulated by CFR. Other than for the periodic removal of oils and sludges in the manner described, there is no other routine maintenance requirements. The plates used in the separator exhibit self-cleaning properties thus routine spray wash cleaning of the plates is unnecessary. During the course of routine operation, the inclined corrugated plate separator will separate sludge, which will settle to the bottom and oil (or fuels) which will rise to the top. The separated oils will continue to collect and the oil layer will increase in thickness unless pumped out or discharged by a floating oil skimmer to an adjacent separated oil storage tank or compartment. As the oil level within this compartment continues to rise, it will eventually fill the separator vessel and/or the waste oil storage tank/compartment. As a matter of routine practice, the separated oil and accumulated sludge should be removed at least once yearly. Of course, if the separated oil storage tank fills more frequently it will have to be emptied more often. The separator should be inspected weekly during the wet season and after every major stormwater event. It is extremely important to have a specific individual responsible for the oil/water separator. This person should also keep a logbook to show the time and date of inspections, sludge removal, oil removal, and all other associated activities.

Flow Bypassing If an oil/water separator module is improperly installed, flow can bypass around the sides, under the bottom, and even over the top. In each such instance a loss of efficiency will occur which could result in an unacceptable effluent. The use of tight fitting modules or flow bypass prevention baffles will prevent bypassing around the sides. The use of sludge baffles significantly reduces bypass under the module. And by extending the height of the modules to the maximum weir head elevation, flow bypassing over the top module can be eliminated. Figure 40 illustrates these requirements.

Figure 40

The Francis Formula can determine the weir head:

Q = 0.415 LH 3/2 2g

where Q = Cubic feet of flow per second Q = Weir length in feet Q = Weir head in feet g = 32.17 ft/sec2

If a separator module becomes plugged bypassing will increase, as water will follow the path of least resistance. Water overflowing the top of a module could indicate a plugged module. It could also indicate that the effluent weir is set at too low an elevation. In either such case the quality of the effluent rapidly disappears as flow bypassing becomes rather nominal. In other words, every effort should be made to minimize the occurrence of flow bypassing.

Fixed Oil Weirs - Use Only With Extreme Caution More often than not water will flow over a fixed oil weir and fill the separated oil storage compartment or dedi-cated vessel. Their use in the marketplace has been consistently unsatisfactory as measured by the number of times that oil storage tanks have had to be pumped or, in the alternative, the number of times that sepa-rated oil has discharged with the treated water. If the rate of flow through an oil/water separator were con-

stant and if the specific gravity of the entrained oil never changed, a fixed oil weir would work satisfactory. Since flow rates and specific gravities frequently change, a fixed oil weir has to be carefully adjusted and frequently monitored to prevent malfunction.

The operation of an oil/water separator with fixed oil and water weirs is hydraulically identical to a U-tube. Since separated oil is lighter than water, a greater tube length is necessary to exactly balance the shorter length of water. As the oil/water separator goes into operation, a weir head develops which, in turn, pushes the oil column upwards by the exact distance of the weir head. In order to understand the rather serious consequences of changed operating conditions, the following examples are examined.

H0 = height from EP to water overflow weir H1 = height from EP to weir head H2 = height from EP to oil weir = oil thickness Figure 41

Example #1 weir head = 1.5” oil weir = water weir + 2” oil sp gr = 0.9 (H1)(water sp gr) = (H2)(oil sp gr) (1) H0 + 1.5” = H1 (2) H0 + 2.0” = H2 (3)

substituting equations (2) and (3) in equation (1) (H0 + 1.5”)(1.0) = (H0 + 2.0”)(0.90) (1) H0 + 1.5” = 0.9 H0 + 1.8” 0.1H0 = 0.3 H0 = 3.0” H1= 4.5” (2) H2 = 5.0” (3) This is the separated oil thick-

ness when weir difference is 2.0”

Example #2 If the oil weir were lowered by 0.25”, the weir difference becomes 1.75” and equations (2) and (3) become:

H0 + 1.5 = H1 (2) H0 + 1.75 = H2 (3) (H0 + 1.5)(1.0) = (H0 + 1.75)(0.90) (1) H0 = 0.75” H1 = 2.25” (2) H2 = 2.50” (3) Oil thickness when weir difference is

1.75”

Note that a difference of only 0.25” in the oil weir re-sults in a 2.50” difference in oil thickness.

Example #3 If the oil weir were raised by 0.25” rather than lowered as in Example #2, equations (2) and (3) become:

H0 + 1.5 = H1 (2) H0 +2.25 = H2 (3) (H0 + 1.5)(1.0) = (H0 + 2.25)(0.9) (1) H0 = 5.25” H1 = 6.75” (2) H2 = 7.50” (3)

Oil thickness when weir difference is 2.25”

A 0.5” adjustment in the oil weir reflected in Examples #2 and #3 results in a total difference in oil thickness of 7.50” - 2.50” or 5.00”. This difference is quite reason-able and should not cause any operational problem.

Two other variables exist which need to be examined. These are oil specific gravity and weir head. If the oil specific gravity charged from 0.90 to 0.85, at the three weir differences of 1.75”, 2.0”, and 2.25” the oil thick-ness becomes 1.67”, 2.20”, and 2.66” respectively.

The oil specific gravity could also increase. If it in-creased from 0.90 to 0.95, at the three weir differences of 1.75”, 2.0”, and 2.25”, the oil thickness becomes 1.00”, 5.00”, and 7.50”, respectively.

In addition to changes in the oil specific gravity, rates of flow to the oil/water separator can also change. These changes can be disastrous. The initial example has a weir head of 1.5”. This can easily change by 0.75” resulting in possible weir heads of 0.75” to 2.25”. With an oil specific gravity of 0.9 and a weir difference of 2.0”, the oil thickness becomes 12.5” and less than 0”, respectively. At oil specific gravities of 0.85 and 0.95, respectively, the oil thickness at a 0.75” weir head are 8.33” and 25.0”, respectively, and at a 2.25” weir head, the oil thickness is again less than 0” in both cases. At this condition water will overflow the oil weir and flood the separated oil storage tank or compart-ment. On the other hand an oil thickness of 25.0” will

definitely prevent significant projected plate separation area from being used and may even flow under the oil dam into the effluent.

From the above analyses one might conclude that by raising the oil weir, water could be prevented from overflowing the oil weir. This happens to be true but an even worse consequence can arise. If the oil weir is raised by 1 full inch it will be 3” above the water weir and 0.5 inches above the maximum weir head of 2.5 inches. This weir head previously resulted in water overflowing the oil weir Now let’s assume an oil spe-cific gravity of 0.85 and a weir head of 2.5”. The calcu-lations on oil thickness become:

H0 + 2.5” = H1 H0 + 3.0” = H2 (H0 + 2.5” (1.0) = (H0 + 3”)(0.85) (1) 15 H0 = 0.05” H0 = 0.33” H1 = 2.83” (2) H2 = 3.33” (3) Oil thickness when weir difference

is 3.0”

At a weir head of 0.75”, the calcs are:

H0 + 0.75 = H1 H0 + 3.0” = H2 (H0 + 0.75”)(1.0) = (H0 + 3.0”)(0.85) (1) 0.15 H0 = 1.80” H0 = 12.0” H1 = 12.75” (2) H2 = 15.0” (3) Oil thickness when weir difference

is 3.0”

Now, however, at an oil specific gravity of 0.95, at the two weir heads of 2.5” and 0.75”, the oil thickness be-comes 10” and 45.0”, respectively.

In summary, low flows and higher specific gravities can translate into extremely thick oil layers which diminish oil/water separator projected area and increase the likelihood of discharging under the oil dam and into the effluent. High flows, even if short term, can cause wa-ter to overflow the oil weir and flood the separated oil storage tank or compartment. Coupled with these op-erational variables, even small vertical adjustments in the oil weir results in magnified differences in the oil thickness. It is not surprising, then, that fixed oil weirs will tend to fail frequently because of variations in flow and oil specific gravity.

A skimming device that works with complete reliability is a floating oil skimmer because:

1. Variations in flow rates (weir heads) cannot ad-versely influence performance because the

floating feature removes this variable. The as-sociated oil weir then rises and falls as the weir head changes.

2. In a well designed floating oil skimmer, the pon-toons or floats rest partly in water and partly in separated oil. As the oil layer increases in thickness, the skimmer sinks slightly because the buoyancy of oil is less than water. The self-compensating feature of a well designed float-ing oil skimmer therefore dampens the varia-tions in oil thickness due to changes in oil spe-cific gravity.

If reliable gravity oil skimming is a design necessity, a floating oil skimmer is the preferred device over a fixed oil weir. A floating oil skimmer may also be deployed as a pump suction inlet thereby permitting a pumped withdrawal of separated oil from the surface of a tank of body of water.

Marketplace Misrepresentation True oil/water separators can be obtained from several suppliers who compete on the basis of equal square feet of effective separation area. Unfortunately, in the last few years, several manufacturers have introduced phony oil/water separators with particular success in the US Government sector. The separators are quite similar in design in that:

1. They use only about 1/10th the required square feet of separation area as determined by basic clarifier hydraulics.

2. The corrugated plates used are oriented axially to the flow rather than transversely. The separated oil then has to tumble upwards from corrugation to corrugation as it migrates to the surface rather than sliding up a smooth trough formed by each corrugation.

3. The use of a “coalescer” pack to “polish” the effluent. These packs are made from closely woven polypropylene (oleophilic or oil loving) netting material, several layers thick. Initially, small suspended solids and entrained oil particles are mostly removed from the flow stream. After a short while, the coalescer pack or element becomes both plugged with suspended solids and saturated with absorbed oil in the same manner as the now obsolete ink blotter would soak up ink. It is quite possible that the proffered phony separator will initially pass the applicable performance tests. The coalescer pack will thereafter plug requiring removal and replacement at considerable cost not to mention the associated requirement to properly dispose of the spent coalescer pack. True separators are mostly maintenance free (other than oil and sludge removal on a periodic basis) devices which last some twenty years plus.

Speaking of test results, be careful to note that vendor sponsored testing invariably fails to show the particle size of the oil droplets removed and instead focuses attention on the oil concentrations of inlet and outlet samples. The entire body of particle separation technology is based upon the capacity to remove discrete particles with specific separation rates. The consideration of concentrations alone is clearly insufficient to show separator performance in the absence of associated data on particle sizes.

Chapters III and V of the API (American Petroleum Institute, the prestigious trade association of the petroleum industry) manual on Disposal of Refinery Wastes as well as API Bulletin No. 1630 are frequently cited as referenced documents in specifications. These citations are incorrect. The classical API Basin is designed to remove all oil particles 150 micron in size and larger whereas plate separators are generally designed to remove all oil particles 60 micron in size. The efficiency of the plate separator is over six times greater than the API Basin. Bulletin No. 1630 merely acknowledges the splendid performance of plate separators generally. It is completely lacking in design information.

Fictitious gravity separators and manufacturer misrepresentations abound in today’s oil/water separator marketplace.

1. One manufacturer uses polypropylene plates and fraudulently claims both sides of each plate in arriving at the total projected plate separation area.

2. The same company incorrectly deploys two and three separator modules in series flow in order to provide the required sq ft of projected or effective separation area.

3. Another successful manufacturer uses vertical plates with inclined ribs. Vertical plates do not project any plate separation area.

4. Another manufacturer also uses vertical plates but with inclined corrugations.

5. Many manufacturers employ polypropylene mesh adsorbers (called coalescers) to compensate for the deficiency of plate separation area. One such manufacturer adds the “surface area” of the polypropylene filaments in the adsorber to the area of the plates in arriving at the total projected plate separation area.

6. Other manufacturers simply misrepresent the square feet of projected plate separation area per cubic foot of media supplied. For example, at a plate spacing of ¾ inch at an inclination angle of 55°, the projected plate separation area calculates out at 9.17 sq ft/cubic foot of media. For other plate spacings and inclination angles see table on page.

7. Several manufacturers incorrectly claim compliance with the API. These manufacturers are engaging in a marketing ploy in an effort to apply API credibility to their product line through name association. The API has not certified a single manufacturer as complying with its design criteria. This may easily be verified by contacting the API.

The above misrepresentations unfortunately inundate the oil/water separator marketplace. Since this equipment is generally sold on the basis of surface area (projected plate separation) considerations, there exists an ever present temptation to overstate a product’s true design. Details of the above designs may be found in WSE Publication 394.

Recommended Specifications The following specifications will insure that the performance requirements will be achieved on free oil removal applications:

1. Fiberglass material of construction for plates - no substitutes will be considered

2. Corrugated plates 3. Cross flow hydraulic path 4. Minimum angle of inclination of 55° 5. Minimum plate spacing of ¾ inch 6. Minimum surface overflow rate of 0.33 GPM/sq

ft (475 GPD/sq ft)

7. Prohibit use of oleophilic materials to aid oil removal

8. Equip separator with sludge collection/sludge removal system

9. Require sq ft of area calculations to be certified by a licensed Professional Engineer

10. One (1) module only in flow path

Laboratory Testing The WaterSmart Environmental laboratory is avail-able to analyze samples for the purposes of determin-ing settling velocities, conducting “jar tests” for floccula-tion or emulsion breaking, and selecting chemical feed dosages. Pilot plant testing is also available.

All laboratory testing is provided on a no cost, no obli-gation basis. Pilot plant testing is available on a nego-tiated cost basis depending on complexity, duration, and location of pilot test site.

The Application Data Sheet following identifies both raw water characteristics, and oil/water separation re-quirements In order to recommend the most cost-effective equip-ment for the application under consideration, it is ex-tremely important to know the requested data. This is especially true when the flow stream contains a multi-plicity of contaminants, only some of which need to be separated to satisfy removal requirements.

Notes

Notes

Application Data Sheet Application: Brief description of oil/water separator intended use:

Capacities:

A. Design Flow Rate:

B. Peak Flow Rate & Duration:

C. Surface Overflow Rate:

Raw Water Characteristics:

A. Inlet Oil Concentration: Of Specific Gravity:

B. Suspended Solids Concentration: Of Specific Gravity:

C. Temperature of Liquid Flow: Normal Maximum Minimum

D. pH of Liquid Flow: Normal Maximum Minimum

E. Oil Particle Distribution, % by Weight

Oil Particle Size in Microns

% Concentration by Weight

F. Concentration of Mechanically Emulsified Oil:

G. Concentration of Chemically Emulsified Oil:

H. Concentration of Dissolved Oil:

Treated Water Requirements: (If NPDES Permit Applies, Attach Copy)

A. Preferred Material of Construction: Steel Concrete

B. Installation is Above Grade Below Grade Yes No

C. Is Oil Storage Required?

Capacity is gallons

D. Skimmer Required?

E. Insulation Required?

F. Immersion Heater(s) Required?

G. Cover(s)?

H. Access Platform(s)?

I. Is Oil Transfer Required?

J. Are Jar Tests or Other Laboratory Analyses Required?

K. Are Pilot Plant Tests Indicated?

No Cost Laboratory Evaluation: Send (freight prepaid) a two gallon sample of raw feed water along with a small quantity of special treatment chemicals, if any, to our Analytical Laboratory. A wide-mouth, non-breakable, and corrosion resistant (plastic) bottle is preferred. Do not send toxic or otherwise dangerous samples without proper identification of the hazardous waste materials.

has been A sample will be sent. Shipping date: via: will not be

Completed Laboratory Report should be sent to:

Company:

Address:

Attention: Whose Phone Number is:

WaterSmart Environmental Local Representative: Notes: 1. The data provided on the Application Data Sheet is our primary source of design information.

The more relevant this information, the faster and more accurate our laboratory work will be. Should we have questions concerning this application, our laboratory staff will contact you.

2. We would like to know a little about the process that generates the oil/water separator application. A process flow sheet would be ideal. We are not interested in receiving any information you regard as confidential or proprietary.

3. If EPA or State permit requirements are to be met, please advise limitations.

Flotation Separators

Application Data Dissolved Air Flotation (DAF) is a process used to remove nearly neutral gravity particles from a flow stream. It does this by using microscopic air (or other gas) bubbles generated in a DAF module. These bubbles become attached to or occluded within the neutral gravity particles, which then float to the surface of the process vessel.

Flotation plate separators have only recently become recognized as another special use of gravity separation technology. This phenomenon is somewhat difficult to understand in that all of the principles of gravity separation still apply regardless of the presence of the DAF process. The use of the DAF process without inclined plates generally results in splendid treatment. The additional use of inclined plates makes the basic process more affordable in a smaller space at higher levels of particle separation.

Although the DAF process has enjoyed wide success in mining, paper & pulp, and chemical processing industries, its overall performance in other areas and applications has not been as good. The DAF process is capable of excellent particle removal performance. To accomplish this high level of treatment requires an appreciation of its inherent limitations.

If the neutral gravity particles are accompanied by a significant quantity of other particles, say 2,000 to 5,000 mg/L range (depending on separation velocities), they must be removed in a gravity type separator prior to the DAF separation system in order to achieve the removal efficiencies inherent to the DAF process. The reason for this kind of primary treatment is twofold.

In the case of the other particles exhibiting rise rates, they indeed will rise to the surface of the vessel under the influence of DAF treatment. However, their utilization of the microscopic bubbles diminished the number of these bubbles available to lift the neutral gravity particles to the surface. It’s strictly a matter of a “numbers” consideration.

In the case of the other particles exhibiting settling rates, again it’s a non-intended utilization of the available bubbles. In addition, the settling rate of some of these particles is reduced because of air attachment resulting in neutral gravity particles that pass through the separator and thus contaminate the effluent.

There are other reasons for apparent failures of the DAF process. In any design, there are only so many bubbles that are available for flotation utilization, namely the excess air saturated flotation flow stream.

In the mining, paper & pulp, and chemical processing industries, the size and number of the neutral gravity particles to be removed by the DAF process have been well identified. Consequently, the amount of air generated in the DAF generation module matches or exceeds the air bubble requirements. Furthermore, these flow streams contain mostly neutral gravity particles thus permitting full use of the available gas bubbles. However, in other applications, the design of the DAF process can fall short for two main reasons.

The use of the air-to-solids ratio, that is, the ratio of the pounds of air available in the DAF generation module divided by the pounds of the suspended solids to be removed, has been relied upon too heavily. The pounds of suspended solids are certainly a good indication of the amount of air that is required to float them to the surface. However, the size of the suspended particles is more determinative of the success of the DAF process than their concentration. If the neutral gravity particles are large, say from 60 microns to 400 microns, the gas bubbles will have little difficulty in finding and attaching to these particles en route to the surface of the liquid. However, if the neutral gravity particles are small, say from 10 microns to 40 microns, the gas bubbles are going to be far less effective in finding and floating these to the surface. Indeed, the air to solids ratio could be the same in both cases, but the relative success far different.

The focus on the part of equipment manufacturers is generating dissolved air rather than on the generation and release of dissolved air. Getting air into solution is only half the answer. The other half is getting the excess dissolved air released. For example, a carbonated beverage doesn’t become “flat” in 2 or 3 minutes, or even in 10. The release of supersaturated gas in a DAF process is no different. To accomplish good release of the excess air requires a good degree of turbulence in the DAF water distribution system. Without this turbulence, similar to shaking the carbonated beverage container, the so-called air available from a design standpoint simply is lost in the effluent discharge. Furthermore, in the design calculations that determine the amount of air available,

to be discounted is that amount of air that is used to merely saturate the DAF flow stream. This air is permanently soluble in the DAF flow stream and cannot be released, even with extreme turbulence. It’s only the excess air, or the supersaturated portion, which is available for release.

The use of the entire stream pressurization, as is popular with many DAF suppliers can lead to poor results. The entire flow stream has to be saturated with air before it even becomes possible to supersaturate. And as indicated prior, most suppliers do not do a good job in releasing the supersaturated portion of the gas. This results in less air and consequently fewer gas bubbles to perform in the DAF process.

When the entire flow stream is saturated and then supersaturated with air, it must all be pressurized. Upon release to atmospheric pressure, the entire stream must pass through a pressure release system. This subjects the entire flow to two successive conditions of particle shearing. The first is the pressure pump that creates the water pressure in the DAF flow. The second is the pressure release valve used to permit pressure relief to gravity conditions. In the shearing of these particles, their numbers are increased and their size reduced both somewhat dramatically. This phenomenon of particle shearing in full flow DAF pressurization is a factor that leads to poorer performance for DAF treatment than is otherwise achievable with this process.

In summary, a comprehensive knowledge of the raw water stream is absolutely necessary for proper application of the DAF process. Additionally, the design of the DAF module is critical to the success of the process. If the particles are too small to be effectively removed by the DAF bubbles, then chemical feed and flocculation will be required for particle growth before air flotation is used.

Table XII

Types of Particles Present

PuriSep Separation System

Neutral Gravity Particles Larger Than 10 Microns DAF Separator

Same As Above Except With Excess Scum Or Solids

Combination 2-Stage Separator/DAF Separator

Neutral Gravity Particles Smaller Than 10 Microns

Combination 2-Stage CorruFloc Flocculator/ DAF Separator

Same As Above Except With Excess Scum Or Solids

Combination 3-Stage Gravity Separator/ Flocculator/DAF Separator

Process Variations All PuriSep DAF Separators behave as gravity separators whenever the DAF system is not in use. It will then separate those particles that exhibit rise and/ or settling rates sufficiently fast to permit simple gravity

separation. Generally, this rate of separation is in the range of 0.75 inches/minute or greater. Should the flow stream vary in process treatment requirements, the above treatment flexibility may be considered.

Sludge Removal Considerations As with all PuriSep Separators, separated sludge must be removed from the sludge holding compart-ment at a rate that matches, or slightly exceeds, the rate of sludge accumulation. All sludges have a ten-dency to harden on storage, and this phenomenon is especially true with chemical type sludges.

DAF Separator Selection The PuriSep Flotation Separator product line has been designed to remove up to 2,000 mg/L neutral gravity particles of a size greater than 10 microns. This design will accommodate the simultaneous presence of up to 1,000 mg/L of suspended solids with settling rates less than 1 inch/minute and/or 2,000 mg/L of particles exhibiting rise rates in excess of 1 inch/minute. Should the concentration of these non-neutral gravity particles exceed the above limitations, they should first be removed in an appropriate separator prior to the DAF process. Should the size range of the neutral gravity particle fall generally less than 10 microns in size, chemical feed followed by flocculation is recommended before DAF removal.

All Flotation Separators are sized at 0.5 gpm/sq ft surface overflow rate disregarding the recycle flow rates. So long as the flow stream under consideration meets the other considerations on the associated particle considerations, it may be applied without restriction.

Laboratory Testing The WaterSmart Environmental laboratory is avail-able to: Analyze samples for the purpose of determin-ing separation rates, determining oil and other particle concentrations and specific gravities, selecting chemi-cal feed dosages, and providing treatment recommen-dations with equipment sizing and budget pricing along with specification preparation.

WSE is also available to conduct pilot plant testing. All laboratory testing is provided on a no cost, no obliga-tion basis. Pilot plant testing is available on a negoti-ated cost basis depending on complexity, duration, and location of pilot test site.

The Application Data Sheet following identifies both Raw water characteristics, and Particle separation re-quirements.

In order to recommend the most cost-effective equip-ment for the application under consideration, it is ex-tremely important to know the requested data. This is especially true when the flow stream contains a multi-plicity of contaminants, only some of which need to be separated to satisfy removal requirements.

Application Data Sheet Application: Brief description of flotation separator intended use:

Capacities:

D. Design Flow Rate:

E. Peak Flow Rate & Duration:

F. Surface Overflow Rate:

Raw Water Characteristics:

A. Suspended Solids Concentration: Of Specific Gravity:

B. Additional Description of Particles Present

C. Temperature of Liquid Flow: Normal Maximum Minimum

D. pH of Liquid Flow: Normal Maximum Minimum

E. Oil Particle Distribution, % by Weight

F. Concentration of Mechanically Emulsified Oil:

G. Concentration of Chemically Emulsified Oil:

H. Concentration of Dissolved Oil:

Disposal method of float material:

Is chemical flocculation required? Yes No Unknown

Treated Water Requirements: (If NPDES Permit Applies, Attach Copy)

Continued on the next page

L. Preferred Material of Construction: Steel Concrete

M. Installation is Above Grade Below Grade Yes No

N. Is Flotation/Scum Storage Required?

Capacity:

O. Float/Scum Transfer Required?

P. Insulation Required?

Q. Immersion Heater(s) Required?

R. Cover(s)?

S. Access Platform(s)?

T. Are Jar Tests or Other Laboratory Analyses Required?

U. Are Pilot Plant Tests Indicated?

No Cost Laboratory Evaluation: Send (freight prepaid) a two gallon sample of raw feed water along with a small quantity of special treatment chemicals, if any, to our Analytical Laboratory. A wide-mouth, non-breakable, and corrosion resistant (plastic) bottle is preferred. Do not send toxic or otherwise dangerous samples without proper identification of the hazardous waste mate-rials.

has been A sample will be sent. Shipping date: via: will not be

Completed Laboratory Report should be sent to:

Company:

Address:

Attention: Whose Phone Number is:

WaterSmart Environmental Local Representative: Notes: 4. The data provided on the Application Data Sheet is our primary source of design information.

The more relevant this information, the faster and more accurate our laboratory work will be. Should we have questions concerning this application, our laboratory staff will contact you.

5. We would like to know a little about the process that generates the oil/water separator application. A process flow sheet would be ideal. We are not interested in receiving any information you regard as confidential or proprietary.

6. If EPA or State permit requirements are to be met, please advise limitations.

191 Standard Conditions of Sale 194 Equipment Maintenance

Wastewater Treatment: Commercial Yards, Garages, and Repair Facilities

195 AquaRound IV™ Car/Truck Wastewater Treatment and Water Reuse System

196 ContamAway II Plus™ Mini Series

198 Capabilities Bulletin 291 Rental and Lease Agreement 294 Stormwater Runoff and

Washdown Treatment: Automotive Dealership/Service Stations

295 Airport Deicing Fluid Treatment and Recovery System

296 SkimAway™ Floating Oil Skimmer

298 RainDrain Plus™ Perimeter Trench Dual Media Filtration System + Phosphorus Removal

380 Silica Contamination Removal from Spent Fuel Pools and Refueling Water Storage Tanks at Nuclear PWR Power Generation Plants

394 A Historical Review of Oil/Water Separator Designs

395 PuriSep™ Differential Gravity Separators

494 Stormwater Runoff and Washdown Treatment: Automotive Wrecking/Salvage

995 OPCT™ - Optimized Physical/Chemical Treatment

996 Schedule of Typical Performance Results: Challenge Oil/Water Separation Testing

998 ABT™ - Aerobic Biological Treatment Process

1094 Mob-to-Demob™ High Performance Aqueous Waste Treatment System

1194 AquaRound II™ Laundromat Wastewater Treatment and Water Recycling System

1195 FilterFresh™ Potable Water Production Plant

1196 ContamAway II Plus™ Mini Series with Spot Free Rinse

1294 AquaRound III™ Laundry Wastewater Treatment and Water Recycling System

1495 EXPRESS™ Simultaneous Ground and Groundwater Remediation

1595 Cost-Effective Energy Savings 1695 UltraPaq™ Packaged Waste-

water Treatment Plants 1795 Selecting an Energy

Management System and Contractor

1895 OAT™ - Optimized Anaerobic Treatment Process

2195 RainDrain™ Perimeter Trench Dual Media Filtration System

498 RipTide™ Pulse Blender/Static Mixer

593 The Removal of PCBs From Aqueous Waste Streams

594 Ansorb™ Adsorbent for Arsenic, Hexavalent Chromium, and Selenium Removal

595 Take a New Look at the RBS Process

598 PuriQuad™ Physical/Chemical Treatment Plant

693 Advanced Aqueous Waste Treatment Concepts

694 OrganoSorb ™ 695 The Biological Approach To

The Rotating Disc Process 791 Quality Assurance Program

Plan 794 DeOiler™ Coalescing Filter

Cartridge 796 Design Manual and Tutorial 894 ContamAway II Plus™

Aqueous Waste Treatment and Water Recycling System

895 Plate Separation - Budding Conventional Technology?

898 Fundamentals of Water and Wastewater Treatment

993 ContamAway I™ High Performance Aqueous Waste Treatment System

994 ContamAway II™ High Performance Aqueous Waste Treatment System

Thank you for your interest in our treatment technologies. To request additional literature simply photocopy this page, mark desired WSE Publication selections, and then mail or fax to WSE with your return address. You may also contact us by phone or email with your request, or visit our webpage. We look forward to helping you find cost-effective and practical answers to your water and wastewater treatment

No. Title No. Title No. Title

Request Additional Information

Post Office Box 26346 Shawnee Mission, Kansas 66225-6346 913.897.2727 fax 897.1902 H2Osmart@ix.netcom.com www.watersmart.com

WaterSmart Environmental, Inc.

5M 3/99 Printed in USA © 1999 WaterSmart Environmental, Inc.

Post Office Box 26346 Shawnee Mission, Kansas 66225-6346 913.897.2727 fax 897.1902 H2Osmart@ix.netcom.com www.watersmart.com