1. Related Applications

This application claims the benefit of co-pending U.S. Patent Application Serial No . 13/372, 182 and entitled CONTROLLED-GRADIENT, ACCELERATED-VAPOR-RECOMPRESSION APPARATUS AND METHOD, which claims the benefit of co-pending U.S. Provisional Patent Application Serial No . 61/443,245 filed February 15, 2011 and entitled CONTROLLED GRADIENT VAPOR RECOMPRESSION SYSTEM and, both of which are incorporated herein by reference in their entireties .

2. The Field of the Invention

This inventio n relates to heat transfer and, more particularly, to novel systems and methods for vapor recompression.

3. The Background Art

Heat recovery is the basis of electrical co-generation plants . Likewise many food and beverage proces ses require heat recovery for economy. Mean while, desalination plants, sugar pro ces sing, distillation systems, and the like rely on recovery of latent heat in order to minimize net energy requirements . Heat may be recovered by reheating, preheating, or otherwise exchanging heat from an exit flow into and incoming flow through a system of heat exchangers .

Vapor recomp re s s io n is used in various forms as one method for heat recovery. For example, in food proces sing, industrial waste proces sing, oil production brine processing, and the like, vapor recompres sion relies on conventional heat exchangers and technologies to exchange heat, vaporize liquids , and condens e dis tillates . The chemical constitution of dis solved materials , especially dis solved solids , as well as various ions and the like take a toll in energy and damage to the processing equipment for energy exchange.

For example, oil production results in pumping considerable water to the surface. That water often contains some amount of hydrocarbons , salts , methane, ammonia, trace elements , or a combination thereof. Therefore the water cannot be released into other waterflows without treatment. Meanwhile, disposal by hauling, followed by re-injection, or evaporation by ponds or boilers , is expensive.

Industrial waste, distillation process in food and beverage industries and the like have similar, if not always so severe, problems . Even the latest methods such as vapor recompres sion and multiple-effect distillation struggle with efficiency, energy budgets, and equipment maintenance in the face of corrosion, fouling, scaling, and so forth. Better systems are needed for heat recovery and re-use.

SUMMARY OF THE INVENTION

In view of the foregoing, in accordance with the invention as embodied and broadly described herein, a method and apparatus are disclosed in one embodiment ofthe present invention as including a controlled gradient of a material, such as, for example, to taldis solved solids (TDS) in a boiling liquid column, such as a brine. Adjacent columns contain condensing vapors at an increased pres sure. High heat transfercoefficients and effective stratified densification ofthe liquid are obtained by controlling mas s flows , work, heat and the like, and sensing and controlling predictively based on balancing mas s, work, energy, and the rates of change thereof, include rates of change in the rates of change (second derivatives ofvalues). In one embodiment ofthe method in accordance with the invention, a system may operate by providing a feed comprising a liquid containing a first material, distinct fromthe liquid and dissolved therein, and containing the feed as a pool. A core at least partially immersed in the pool may be in thermal communication therewith and sealed against direct fluid communication therewith.

One may create a concentration pro file reflecting a variation ofthe concentration ofthe first material in the pool between a liquid level at the top thereofand the bottom thereof by recycling vapor produced in the pool into a condensate within the core. Typically, a container containing the feed is selected from a pond, a tank, an estuary, and a ves sel, and the pool is quiescent relative to the feed.

The core may further comprise closed channels in thermal communic atio n with the pool, in indirect fluid communication therewith, and sealed against direct fluid communication therewith, which may be oriented to flow the vapor and condensate in a vertical direction. Controlling accretion of compositions containing the first materialmay be done by selecting the attitude ofthe core in operation.

The portion of the pool within the core may be engaged in confined boiling, and the profile (which may be thought of as gradient, but is not neces sarily monotonic orlinear) is effected by establishing an exchange of heat from the core into the pool. A change in phase ofthe liquid within the core, by confinement therein vaporizes the liquid during the heat transfer from the core.

Optimizing the concentration pro file may be done by providing a plurality of panels and selecting a spacing therebetween for enclosing therebetween, in at least two dimensions , a portion ofthe pool. For example, this may include providing a plurality of panels and selecting a spacing therebetween for enclosing therebetween, in at least two dimensions, a portion ofthe pool. Spacing may be based on the characteristics ofthe feed.

The method may include selecting at least one of a spacing between panels ofthe plurality of panels, a number ofthe panels in the core, a size ofthe panels , materialofthe core, attitude ofthe core, othercharacteris tics ofthe panels, and the position ofthe core in the pool, and a combination thereof based on the characteristics ofthe feed.

Operation ofthe systemandmethodestablishes an active region proximate the core and containing a substantial majority of the variation in the concentration profile, and establishes a trap region below the active region, which is substantially excluded from exchanging liquid into the active region.

Optimizing heat transfer may be done by fully immersing the core into the pool, and controlling or changing an effective nucleate boiling region ofthe core by changing the concentration pro file. Changing a temperature profile in the pool by adding heat corresponding to a change in a pres sure above the pool may be done, and may be balanced with work by the compres sor to obtain stability at a set of conditions desired.

In one embodiment ofamethodinaccordance with the invention, a proces s may include changing a temperature profile in the first region by adding heat based on a change in a pres sure above the first region. Changing a boiling region ofthe core may be effected by changing the concentration pro file, which may be used to change the effective saturation temperature, pres sure, or both for the liquid. The pool may be quiescent relative to the feed, meaning that flows are generally comparatively slower, with turbulence only local, and not general.

One embodiment of an apparatus in accordance with the invention, may include a containment means adapted for receiving a feed comprising a liquid containing a first material distinct from the liquid and dis solved therein. The containment means may be configured to contain a collection ofthe feeds as a pool having a liquid level and a bottom level. A core may be at least partially immersed in the pool to be in thermal communication therewith and sealed against direct fluid communication therewith. Means for proces sing the pool may create a concentration profile reflecting a variation in concentration of the first material in the pool between the liquid level and the bottom. This proces sing means may further comprise compres sion means recycling vapor produced in the pool into a condensate within the core, and may include heating means (such as a heater, for example) for adding thermal energy into the pool. The processing means may include a compres sor, which is one embodiment of a recycling means for recycling vapor produced in the pool into a condensate within the core.

Containment means may be selected from a pond, a tank, an estuary, a vessel, or the like. The core may include closed channels in thermal communication with the pool, in indirect fluid communication therewith (e.g. to receive vapor), and sealed against direct fluid communication therewith. The core may be movable, for moving relative to the containment means . When the core is engaged in confined boiling, moving may be used to adjust spacing between panels ofthecore. Moving the core may include changing the orientation ofit, changing a spacing between the closed channels, or the like.

In one embodiment of an apparatus in accordance with the invention, configured as a heat exchanger suitable for use in a medium configured as a fluid, the heat exchanger may include an inlet, outlet, and surfaces . Surfaces may include an exterior surface and an interior surface, defining an interior volume in fluid communication with the inlet and the outlet.

The surfaces may be constructed ofa material selected to have a thermal resistance for optimizing heat transfer from the interior volume into the medium (fluid). The material's properties considered may include a coefficient of thermal expansion effective to maintain the geometric structural integrity of the surfaces, effective to be stable in an environment comprising the medium, effective to minimize nucleation during boiling of the medium thereagainst, or a combination thereof.

The inlet may conduct a recycled vapor, generated against the exterior surface, into the interior volume, the exterior surface conducting heat from the interior volume into a boundary layer formed by the exterior surface when contacted by the medium. For example, the material may be selected from the group consisting of metals , polymers , composites , and a combination thereof. One suitable polymer is a fluoro carbon polymer, such as a tetrafluoro ethylene (e.g. polytetrafluoroethylene).

The material may selected to be chemically inert and non-reactive with respect to the medium. It may also be selected to minimize accretion of compounds generated in the medium.

A method for improving a proces s for vapor recompres sion, may include selecting a proces s comprising a plurality of operations combinable as sub units to effect the proces s . Determining a concentration profile ofa material dis solved in a source of the vapor may be done in conjunction with determining an influence on the concentration profile. This may be done b y ev aluating at least one operation of the plurality of operations having a set of operational parameters .

Selecting a target operation fromthepluralityofoperations may be based on that evaluating. Selecting a control parameter for controlling the target operation, one may begin manipulating the concentration profile by modifying the control parameter. The control parameter may be selected from the group consisting of a mas s flow, mechanical work, thermal energy, thermal inertia, a rate of change thereof, and a combination thereof for certain embodiments . In other embodiments a larger group may be considered Evaluating may consist of evaluating in sequence a pump moving liquids in the proc e s s , a c o mp re s sor compres sing the vapor from the source, and a heater adding heat to the source. It need not include more than those actions, but could include evaluating the response time ofthe source (e.g., thermal inertia).

In one embodiment, evaluating may also be sequentially and in an order of first, a pump for moving liquids in the proces s , second, a compress or for compres sing the vapor from the source, and third, a heater for adding heat to the source. These may be evaluated when actually moving liquids , compressing the vaporfromthe source, and adding heat to the source.

The method may include analyzing the feed for at least one of the constituents therein, time variance ofthe constituents , a source of supply, delivery mechanisms , and the like. The method may include modifying a control corresponding to at least one of a pump, a compres sor, a heater, and a combination thereof. It may also include providing sensors to detect at least one of a temperature, pres sure, flow rate, power, and concentration corresponding to an operation within the proces s . It may beneficially include determining an ambient condition selected from pressure, temperature, wind, humidity, and a combination thereof.

Evaluating may include determining substantially all (or all) energy inputs into and energy outputs from the proces s . The method may thus include balancing substantially all inputs of energy into and outputs of energy fromthe proces s . It may add an energy recovery operation providing energy transfer with respect to at leas tone ofthe operations .

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and otherobjects and features ofthe present invention willbecome more fully apparent fromthe following description and appended claims , taken in conjunction with the accompanying drawings . Understanding that these drawings depict only typical embodiments ofthe invention and are, therefore, not to be considered limiting of its scope, the invention willbe described with additionalspecificity and detail through use ofthe accompanying drawings in which:

Figure 1 is a schematic block diagram of a controlled-gradient, vapor-recompression systemin accordance with the invention;

Figure 2A is a schematic block diagram ofthe system of Figure 1 implementing sensors and a controller for operation of key apparatus and parameters ;

Figure 2B is a schematic block diagram of the control system, illustrating the inner-most and outer-most levels of control;

Figure 2C is a schematic diagram of a mechanism for detecting liquid level in a tank in accordance with the invention, without interference from turbulent surface activity;

Figure 3 is a schematic diagramofvarious embodiments ofoption configurations of components for the system ofFigures 1-2;

Figure 4 is a schematic diagram of a core within a tank in accordance with the apparatus and method ofFigure 1, illustrating the activity ofthe heat and brine convection processes ;

Figure 5 is a chart illustrating the relationship between total dis solved solids in a tank ofthe systemof Figure 1, between the liquid level and the outlet level ofthe tank;

Figure 6 is a chart of curves illustrating the normalized total dis solved solids increase in the concentration gradient or density gradient ofthe tank of Figure 5, in a system of Figure 1 ; Figure 7A is a chart of the temperature as a function of height in the tank of Figure 5, equipped with a core of Figure 4 in the system of Figure 1 ;

Figure 7B is a description o f Raoult's Law governing saturation temperature in impure liquids , such as production brine;

Figure 7C is a description of the Clausius -Clapeyron equation describing the change of temperature in a vapor across a compres sor increasing the pres sure on that vapor;

Figure 7D is a description of Dalton's Law of partial pressures in a ves sel containing multiple gas ses ;

Figure 7E is a description ofHenry's Lawgoverning the concentration of absorbed non-condensable gas ses as a function of pressure contribution of those gas ses above a liquid in equilibrium;

Figure 8 is a perspective view of one embodiment of an apparatus in accordance with Figures 1-7;

Figure 9 is a table representing input variables into an experiment in which brine is concentrated from an initial feed waterconcentration level to an output brine concentration levelin an apparatus and method in accordance with the invention;

Figure 10 is a chart illustrating curves representing the concentration or density gradient change in the experiments outlined by Figure 9 and implemented in the system of Figure 8, showing the normalized total dis solved solids increase as a function of liquid level in the tank of the system of Figures 1-9;

Figure 11 a chart showing the temperature of saturation in the experiment of Figures 9-10, and compared with the expected performance of a conventional, completely mixed heat exchange system; and

Figure 12 is a schematic block diagram of a method of controlling the system of Figures 1-11, in accordance with the levels of control illustrated in Figures 2A-2B.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

It will be readily understood that the components of the present invention, as generally described and illustrated in the drawings herein, could be arranged and designed in a wide variety of different configurations . Thus , the following more detailed description ofthe embodiments ofthe systemand method ofthe present invention, as represented in the drawings, is not intended to limit the scope ofthe invention, but is merely representative of various embodiments ofthe invention. The illustrated embodiments ofthe invention will be best understood by reference to the drawings , wherein like parts are designated by like numerals throughout.

As used herein, terms are to be understood and interpreted broadly. However, alternative, specific terms may be used by way of example, but are to be interpreted as meaning the broader terms . For example a solvent or liquid is exemplified by water, but may be interpreted as any s o lvent, liquid, material, medium, carrier, or the like. Similarly, many materials may be dissolved as solutes in such a carrier. Solutes may be called contaminants herein; contaminant simply refers to something to be separated out, even a desirable material as in distillation. A solvent or liquid may be thought of as any fluid to be treated by a separation proces s in accordance with the invention.

Solutes may be liquids , solids , ions , synthetic, natural, mineral, animal, vegetable, or other materials dis solved in the solvent or carrier. Thus , the term TDS is an example standing for a solute generally, dissolved in the carrier as its solvent. Solutes and solvents may arise in foodprocessing,industrialproces s fluids orwastewater,alcoholdistilleries , sugar proces sing, petroleum drilling or production fluids , potable water proces sing, mining effluent or tailings proces sing, nuclear coolant or waste liquids proces sing, runoff or other collection pond handling, or the like. Brine stands for any solution of solute in a solvent, even though it is an example term commonly applied to dis solved solids and ions in water.

Core materials may be any suitable materials ranging through metals , alloys , stainles s , polymers , elastomers , othermaterials , composites, orcombinations thereof. The core may have bellows structures to change spacing between panels , or other variations supporting positioning, pivoting, tilting (e.g., attitude of roll, pitch, or yaw around any axis), sliding, orotherwise optimizing configurations of core panels by positioning. Such may be usefulin proces ses such as vapor recompres sion, evaporator distillation systems , multiple-effect evaporators , and otherprocessing systems , even though otherwise difficult in some industrial situations .

The core described herein is not a 'radiator' like an automobile uses, for several reasons . For example, air through such a radiator is a flow completely unrelated to the cooled liquid contained. In contrast, vaporrecompression pas ses a vaporphaseboiledoffa liquid phase, through a compres s or and back to condense against the out sideofthe very wall containing the boiling.

Quiescent is comparative between flows , and does not mean a complete lack of flow or motion, but rather a much slower motion than the flow compared with it and conventional flows for the function. Nucleate boiling is not limited to boiling initiated at surface nucleation points , butboiling due to exceeding the vaporpres sure. Confined boiling is a term of art in the art of heat transfer and is used in its ordinary meaning therein. It is also understood to mean nucleate boiling in a space confined in at least one dimension.

Likewise, fluids include allgases , vapors , liquids, and liquidous flows . Systems or devices in thermal or fluid communication mean the systems are capable of exchanging heat or fluid, respectively. Containment for fluids may include anything natural or artificial, from ponds , lakes , rivers , and other estuaries to lined ponds , tanks, containers , pipes , conduits , or the like.

By gradient is meant a profile (a variation in one variable, like temperature or concentration, with respect to another, like space or time). It need not be linear, nor monotonic (changing always in a single direction). The profiles often tend in one direction, with localized variation due to the dynamics ofthe system. Typically, apro file changes more dramatically in an active region (region where heat transport, mas s transport, or both are actively occurring between flows , and not just flowing through some containment from one location to another). A dynamic gradient or dynamic profile is a profile established by operation of the invention, and subject to localized variations , variations with time or conditions , or a combination thereof.

Referring to Figure 1, while referring generally to Figures 1-12, a system 10 in accordance with the invention may be set in a permanent installation, or may be containerized. The basic elements of system 10 may include a tank 30.

In the illustrated embodiment, the tank 30 contains a brine 23 that has established therein a gradient of concentration of the dis solved solids . The tank 30 is fed originally by a feed tank 32 through lines 33. In general, herein, any reference to an itemby reference numeral includes a generalized inclusion of such items bearing such a number. A trailing letter after a reference numeral indicates a specific instance of the item designated by the reference numeral. Thus , the system 10 includes a plurality of lines 33, including, for example, lines 33a, 33b, 33c, and so forth.

The feed tank 32 provides through lines 33 to a separator 34 a flow 35. The flow 35 is typically pre-treated in the separator 34. In o ne embodiment, the separator 34 may be configured as a pre-treatment system for removal of volatile materials, for example.

In one embodiment of a method in accordance with the invention, the system 10 may be used by introducing a brine 23 in an unconcentrated state into the feed tank 32. This may come directly from a well head, or may be hauled to a particular location from various petroleum production facilities . In the illustrated embodiment, the feed tank 32 may then transport the brine through a line 33 to apre-treatment system 34, which typically will operate as volatiles separator 34. Other processes of pre-treatment systems 34 may include adding various chemicals in orderto reduce fouling, scale, corrosion, and the like.

For example, brine received in a feed tank 32 may include numerous materials . Dispersed oil products are typically volatiles that vaporize upon heating. These may include fractions ofcrude oil that range fromC6 to waxes , tar, paraffin, as well as paraffin soluble organic compounds . Gasoline and diesel ranges of organic hydrocarbons may be included in small amounts . Likewise, various aromatics , such as polycyclic aromatic compounds may be included. BTEX compounds are not uncommon. Likewise, methanol, phenols , and methane may similarly be included.

Not only those organic hydro carbons butlikewisesulfurin various forms , including hydrogen sulfide (H ₂ S) may be included. These may be particularly problematic since sulfates are likely to permanently scale out on solid surfaces . In orderto reduce the scaling by sulfates , scale inhibitors must be introduced into the brine 23 to maintain a clean feed tank 30. These are not necessarily required, but are highly recommended forbrines 23 that contain compounds ofsulfur.

Similarly, silica, clay, and other inorganic materials may be included in large or small amounts, dissolved, or undis solved. Typically, silica and clay are undissolved, and may form particulates . Likewise, various salts . Salts may include cations ranging through magnesium, calcium, sodium, and potassium. The anions , which may correspond to the aforementioned cations may include chlorides , sulfates, carbonates, nitrates , and the like. Typically, nitrates are not present in large concentrations . Nevertheles s, carbonates are typically received in brines 23 in comparatively large or larger quantities .

Treatment chemicals added in the pre-treatment system 34 may include, for example, ammonium, various compounds of nitrogen, gels , foam generating materials , and the like. Similarly, additional ions may include strontium, mercury, lead, chromium, selenium, iron, barium, and so forth. Various naturally occurring radioactive materials such as uranium, radium, and the like may be included. Boron is not all that uncommon.

In some emb o diments , various types of separators 34 may be placed to remove other entrained materials, whether solid, gas , liquid, or the like. Such pre-treatment systems 34 are numerous and ubiquitous in the science of pre- treating production brines .

For example, Sears , in United States Patent No . 5,968,321, is sued October 19, 1999 and entitled Vap o r Compression Distillation System and Method, which is incorporated herein by reference, discloses a distillation system that includes a pre-treatment proces s and apparatus . Similarly, Kresnyak, et al., in United States Patent No . 6,355,145 Bl is sued March 12, 2002 and entitled Distillation Process with Reduced Fouling, which is incorporated herein by reference, likewise discusses various proces ses for pre-treatment.

From the pre-treatment system34 or separator 34, the flow 35 first pas ses through a heat exchanger 36, referred to as a brine heat exchanger. The function of the brine heat exchanger is to remove heat from brine 23 leaving the tank 30, and to recover that heat into the flow 35 pas sing into the tank 30.

Ultimately, the concentrated brine from which heat is extracted by the brine heat exchanger 36 is disposed of in a brine tank 38. The brine tank 38 may be emptied by hauling the brine away, pas sing the brine into an evaporation pond, further proces sing the brine for minerals , heating or otherwise drying the brine, or other disposition method.

In the illustrated embodiment, a distillate handling system40 operates opposite the brine heat exchanger 36 and brine tank 38. That is , forexample, the distillate handling systemreceives the distilled water as an output fromthe system 10, and specifically from the tank 30 where it has been boiled off. The distillate handling system 40 may include a vapor trap 41. The vapor trap 41may be simple or complex and typically operates like a liquid trap (e.g., P trap) in which a column of liquid is contained within a line 33 that traverses both down and back up in order to maintain a liquid column that cannot be overcome by the pres sure of incoming vapor.

The distillate tank42 operates to collect all the distillate that has been condensed from the closed channels 24 of the core 20. However, as a practical matter, particularly in consideration of control is sues , a distillate reservoir 43 may first receive the distillate fromthe vapor trap 41. Accordingly, the distillate reservoir 43 may be used for testing the level or rate of generation of distillate.

Following collection in the distillate reservoir 43, the distillate may next pas s to a heat exchanger 44 configured to extract heat fromthe distillate, and pas s that heat into the feed input line 33a feeding into the tank 30. In the illustrated embodiment, the distillate heat exchanger may operate at a fixed rate of flow in both directions .

For example, the brine feeding from the feed tank 32 may be divided between feeds F l, passing through the distillate heat exchanger 44, and F2, pas sing through the brine heat exchanger 36. Thus , F l receive s heat from the distillate, preheating as close as reasonable to the temperature ofthe brine 23 in the tank 30. Likewise, feed passing from the feed tank 32 through the brine heat exchanger 36 extracts heat from brine exiting at maximum concentration from the tank, toward the brine tank 38. This preheating of Fl and F2 elevates feed temperatures and recovers heat that would otherwise be discharged in the distillate tank 42 and the Brine Tank 38, respectively.

In the illustrated embodiment, the distillate handling system 40 includes a level control 45. The level control 45 operates by sensing the level of distillate in the reservoir 43. According to the output of the level control 45, the system 10 may be adjusted in certain operating parameters in order to maintain a constant flow of distillate.

In the embodiment of the illustration, it is contemplated that the distillate outflow to the distillate tank 42 from the distillate handling system 40 through the distillate heat recovery system 47, will be operated at a fixed rate. One benefit of an apparatus and method in accordance with the invention is that the output rate of distillate may be fixed. Likewise, the mcoming brine mas s flow rate may be fixed in the flow 35, divided between the flows F l, F2, regardles s ofthe brine concentration incoming from the feed tank 32, and regardles s of the brine concentration level discharged into the brine tank 38.

Various embodiments of level controls 45 may be implemented. For example, Figure 2C hereinafter describes one level control mechanism suitable for operating between a vapor compartment and a liquid compartment or a vapor region and a liquid region within a tank, while still providing accurate, repeatable, reliable readings, without the need for vents and other condensate removal systems from the vapor side of the gauge.

The energy sources for evaporation of the brine 23 in the tank 30 comes from multiple sources . As a practical matter, an auxiliary heat source 46 provides heat to supply the brine 23 in order to elevate the temperature within the tank 30 to the proper level. Meanwhile, the brine heat exchanger 36 and distillate heat exchanger 44 recover heat from exit streams in order to elevate the temperatures of F2 and F l, respectively, entering the tank 30.

Thus , the distillate heat recovery system 47 is a source of heat recovered into the line 33a, as the brine heat recovery system 80 is a source of recovered heat into the flow 35c in the line 33a. Other heat recovery systems such as engine exhaust recovery may also be employed. Actualsources of heat will typically include only a heater 70 providing heat from an auxiliary source 46, which operates merely to overcome los ses in the system.

The tank 30, may include a level control 48, which may be similar, or completely different from the level control 45 on the distillate reservoir 43. Each of these level controls 45, 48 may operate substantially independent ofthe rest of the system 10. However, in certain embodiments, the level controls 45, 48 may operate directly to control the feed 35a through the lines 33a, in order to match mas s flow rates according to conservation of mas s .

An ancillary option at an appropriate place in the system 10 may be a distillation column 49. It has been found useful in some production water sources to implement a distillation column 49 in order to remove heavier materials, such as distilled water, in a stripping section, while separating out lighter components, such as methanol or the like, in a rectifying section at the top thereof. Thus , the distillation column 49 is an optional element that may or may not be included depending upon the particular site being serviced by a system 10.

A compres sor 50 compres ses vapor 27 originating in the tank 30 in the brine 23, and collecting above the brine 23. The compres sor 50 is responsible to raise the pressure in the vapor 27 according to the Clausius -Clapeyron equation relating temperature rise to pres sure rise. Accordingly, the vapor 27 pas ses through the compres sor 50 and is fed back into the manifold 19 ofthe core 20.

The pressure downstream of the compres sor 50 exists substantially the same in the conduit l8, manifold 19a, and the close channel 24. The differential in pres sure between the upstream side of the compressor 50 and the downstream side thereof effects a pressure of saturation corresponding to a higher temperature of saturation.

Heat is transferred due to the temperature differential between the closed channels 24, of compres sed vapor, and the open channels 22, in the brine 23. Heat from the condensing, saturated vapo r 27 in the closed channel 24, transfers into the brine 23.

In some embodiments , a vapor handling system 52 may be mounted near or at the top of the tank 30. In the illustrated embodiment, the vapor handling system 52 may include, for example, a mist eliminator 54. Typically, a mist eliminator 54 is responsible to remove droplets of water, which may entrain droplets of brine 23, from the vapor 27 and bring with them the risk of carrying dis solved solids toward the compres sor 50.

Various embodiments of vapor handling systems 52 may be considered. In addition to the mist eliminator 54, for example, a deaerator 56 may be included as part ofthe vapor handling system 52. De-aerators at this stage need not be exces sively large, nor vent substantial quantities ofthe vapor 27.

For example, in one apparatus and method c o n s tructed for experiments , and producing approximately 100 barrels per day of distillate in the distillate tank 42, a de-aerator 56 was sized by conventional chemical engineering principles . Henry 's Law, which relates concentrations of non-condensables or othervapors within liquids , according to the partial pres sure and a physical constant, as described hereinbelow, required a reservoir of about twenty liters . Accordingly, the deaerator 56 needed only about five liters to be vented approximately once per day during operation.

In general, a plenum 58 above the brine 23 in the tank 30 may be sized to provide a dwell time or accumulation time for v apors 27 in order to enhance mist elimination. Numerous manufactures produce compressors of constant displacement, positive displacement, and so forth. For example, Ingersoll Rand, Dresser, and other companies produce compres sors 50 suitable for application in a system 10 in accordance with the invention. Likewise, a plenum 58 may be sized according to the rating of a compressor 50.

Ultimately, the brine 23 is co ncentrated by boiling and vaporizing the brine 23 into vapor 27. As vapor 27 leaves the brine 23 as bubbles , and enters the plenum 58, the residual dissolved solids within the brine 23 increase in the region about the vaporized bubble. This increase in dissolved solids in this surrounding brine 23 results in higher density and a net downward flow of this more dense brine 23.

Ultimately, the tank 30 establishes a c o nc entration profile or gradient, in which the brine 23 of lowest concentration exists at the interface between the brine 23 and the vapor 27. A ccordingly, the heaviest or the most concentrated brine 23 is established at the output level of the tank 30. The function of the system 10 is to concentrate brine 23 from whatever concentration exists in the feed tank 32 to a much greater concentration.

As the brine 23 loses water into vapor 27 collected in the plenum 58, localized concentrating processes occur around every bubble formed. These localized concentrations , result in localized descent of heavier brine 23 relative to lighter brine.

For example, the brine 23 in the feed tank 32 has les s dis solved solids , and is lighter, per cubic inch or cubic centimeter than the brine in the brine tank 38. In the locality of bubbles , a density differential develops beside a bubble that has vaporized. The bubble leaves behind its share of dis solved solids to be absorbed by neighboring liquid water molecules in the brine 23. Ultimately, with the continuing proces s of heating and evaporation occurring within the open channels 22 as a result of the heat transferred from the closed channels 24, a continuing concentrating proces s occurs within each open channel 22.

As a direct result, heavier, more concentrated brine 23 moves downward seeking density equilibrium among equally agitated boiling or near boiling neighbors . Thus, in steady state the maximum concentration of dis solved solids exists at the outlet of the tank 30 and the minimumdensity and minimum concentration of dis solved solids exists at the interface between the brine 23 and plenum 27. This has been demonstrated in experiments .

A concentrate handling system 60 is responsible for handling the concentrated brine 23 exiting the tank 30. In the illustrated embodiment, the concentrate handling system60 includes a slurry handling system 62. The slurry handling system 62 is responsible for handling such items as high density precipitates that may form sludge, or other suspended solids at high concentrations in liquid. Accordingly, such materials may be separated from the brine 23 of the tank 30 and directed to disposition different from the brine in the brine tank 38.

Similarly, a reservoir 64 may act as a settling tank 64, as well as concentrator 64. As a practical matter, concentrations , having orcausing the greatest stratification, occurin regions where concentrating activity, such as boiling evaporation are found. In the illustrated embodiment, that region is the region within the open channels 22. In contrast, the reservoir 64, lacking any heating or evaporatio n mechanism concentrate may operate as a settling region, and typically does not concentrate substantially further.

Likewise, the brine concentration system 60 may include a variety of mechanisms within the slurry handling system 62 to as sist in removing precipitates and other solids from the walls , floors , and the like of various components .

Solids removal equipment is known in the art and may include vibration systems , scraping systems , augers , combinations thereof, and the like. Ultimately, a slurry holding system 66 may actually be separated by valving fromthe slurry handling system 62, and only receive brief and periodic discharges ofsolidus flows into the slurry holding system 66. Such systems may be manual, automatic.

Following pas sage through the pre-treatment system 34, the brine 23 may pas s through a flow divider 74, such as a valve or system of valves dividing the overall flow in the lines 33 from the feed tank 32 into F l and F2, illustrated by feed 35a pas sing through line 33a. The flow divider 74 is responsible for maintaining a constant flow to the distillate heat recovery system 47, and a variable flow to the brine heat exchanger 36. The control of the relative proportion of these flows will be discus sed hereinbelow.

In the illustrated embodiment, the flow 35a through the lines 33arepresents two flows . A fixed rate through the distillate heat exchanger 44 is matched to the fixed flow of the distillate through the distillate heat exchanger 44.

In contrast, the fresh brine from the feed tank 32 pas sing through the brine heat exchanger 36 is adjustable, commensurate with the flow of brine concentrate out of the tank 30, through the line 33c into the brine heat exchanger 36. In all cases , metered pumps , controlled typically by being fixed displacement pumps 76, may be placed in the lines 33 to control the rates of the flows 35 into and out of the tank 30. For example, feed pump 76a may control the flow of brine from the feed tank 32 through the distillate heat exchanger 44.

Likewise the pump 76b controls F2, or the flow 35a pas sing from the brine heat exchanger 36 into the diffuser 68 in the tank 30. Similarly, a brine pump 76d may control the feed of the concentrated brine from the tank 30 through the brine heat exchanger 36. A pump 76c may controlthe flow of distillate into the distillate heat exchanger 44, and may be matched by mass flow of the brine pump 76a.

Continuing to refer to Figure 1, while referring generally to Figures 1-12, a system 10 in accordance with the invention typically feeds preheated brine 23 through the lines 33a into a diffuser 68. The diffuser 68 is described in various options in Figure 3. By whichever mechanism is selected, the diffuser 68 has the effect of distributing the brine 33a, in a fashion that will allow the brine 23 from the lines 33a to distribute across the maximum extent of the core 20.

For example, the cross -sectional area or foot print of the core 20 in the tank 30 represents a particular area of interest. That area p re s ents the b o ttom openings of all the open channels 22. Accordingly, a diffuser 68 may be responsible to distribute the flows 35a of the lines 33a within the core 20. If all the flow 35a pas ses into a single open channel 22, the efficiency of the core 27 will be different from that achieved if all open channels 22 of core 20 have a reasonably equal opportunity to receive a portion of the flow 35a.

An auxiliary heater 70 is responsible to add heat received from a heat source 46 or an auxiliary heat source 46. In the illustrated embodiment, the auxiliary heater 70 is positioned below the bottom of the core 20.

In certain embodiments , the auxiliary heater 70 may be placed on a wall 31 of the tank, rather than inside the tank 30. Likewise, the auxiliary heater 70 may be res p ectively positioned with respect to the diffuser 68, such that auxiliary heat source 46 feeds heat directly into concentrated brine at the bottom of the tank 70, rather than into the incoming brine 35a flowing into the diffuser 68.

In some embodiments , a diffuser 68 may not be required. In others , an engineering selection may be made between heating the incoming brine 35a with the auxiliary heater 70, and allowing the incoming brine flow 35a to simply rise due to saline convection (TDS content convection) to the top of the core 20 without the benefit of carrying any heat.

In Figure 1, diffuser 68 is positioned at a level below the auxiliary heater 70. Thus, the flow 35a from the input lines 33a emitting from the diffuser pass through the layer of heated brine created by the auxiliary heater 70. This provides a heat transfer mechanism for heating the tank brine 23. In some embodiments, the auxiliary heater may actually be located in the diffuser. In other embodiments , the auxiliary heater 70 may be attached to the inside or outside of a wall 31 of the tank 30. In otherembodiments , the auxiliary heater 70 may actually be in the lines33a feeding into the tank 30.

In order to monitor, and subsequently control operation of, the system 10, sensors 72 may be installed in the system 10. Sensors 72 may include sensors 72 to monitor pres sure, temperature, concentration of dis solved solids , combinations thereof, or the like. In the system 10, concentration effectively improve heat transfer, and mass transfer (evaporation and condensation, for example) by virtue of even small differences in concentration. Hence, temperature, pres sure, and concentration measures are significant as control parameters in the brine 23 and the vapor 27. Control of system 10 may require a multiplicity of these sensors 72.

Nevertheless , with such items as the compres s or 50, lines 33, conduits 18, and other fixtures , pres sures may vary throughout the system 10. Mean while, inas much as the system 10 operates about saturation pres sures and temperatures , temperature is an indicator ofp res sure, and vice vers a. Thus , each may be sensed, and steps may be taken to assert active control in accordance with established functional relationships . Concentration profiles, which may be referred to as gradients , of dis solved solids are established within the brine 23 ofthe tank 30, and thus localized density may be implied by those concentrations . Accordingly, density changes , altitude changes , together with any pres sure changes within the plenum 58, may add up to provide a comparatively wider variety of pressure and saturation temperature variations at points throughout the tank 30 than would a mixed tank 30. Thus , monitors and control systems may be in place to read the sensors 72 and feed that data to actuation devices .

In the illustrated embodiment, sensors 72a are positioned within the open channel 22 exposed to the free stream or bulk of the tank 30. Sensors 72b are located within the closed channel 27. Sensors 72c detect conditions within the tank 30 near the wall 31. The sensors 72c may be placed at the wall, but will more typically be placed in the brine 23 spaced from the wall 31, but mounted to the wall 31. Sensors 72d exist in the plenum 58 to detect conditions therein.

Likewise, sensors 72e in the vapor handling system 52 detect conditions therein, while sensors 72f monitor the heaviest brine 23 concentrated at the bottom of the tank 30. The region ho sting the sensors 72f does not have any portion ofthe core 20 active therein but may be important in the control of system 10.

The compres sor 50 may be monitored by sensors 72g on the upstream or inlet side thereof, and sensors 72h on the downstream or outlet side thereof. Ambient conditions may be monitored by sensors 72j external to the tank 30, located in the environment to sense ambient and atmospheric conditions .

From the plenum 58, the conduits 18 carry the vapor 27 into the compres sor 50, and from the compres sor 50 into the plenum 19a ofthe closed panels 24 or closed channels 24. The vapor 27 within the closed channel 24 eventually condenses to form the condensate 25 in the bottom of the closed channel 24. Eventually, the lower plenum 19b ofthe closed channels 24 may be completely filled with liquid.

Nevertheless , it may be pos sible that some vapor 27 may be circulated through the dis tillate 25 or the condensate 25 at the bottom of the closed channels 24. Accordingly, the flow from the closed channel 24 into the vapor trap 41 may contain both gas and liquid phases ofthe condensate 25.

Meanwhile, the levelcontrol45 monitors the levelof condensate 25 in the reservoir43. Ultimately, the reservoir 43, controlled by the pump 76c pas ses the distillate through the distillate heat exchanger 44 and on to the distillate tank 42.

Referring to Figure 2A, while continuing to refer generally to Figures 1-12, the system 10 may include a controller 84. In general, controller 84 includes at least one pro ces s or, and typically the complete input systems , output systems , processing facility, memory, and so forth of a computer. The controller 84 may receive data, proces s data, store data, and so forth. The controller 84 is responsible to receive inputs from sensors 72 throughout the system 10.

Specifically, the controller 84 will receive information in the form of data regarding temperatures, pres sures , concentrations , and so forth as wellas flowrates , and the like from the various components described hereinabove with respect to the system 10. In the illustrated embodiment, the controller 84, although illustrated multiple times, may be a single proces sor-based system, or multiple proces sors . The controller 84 may be consolidated, distributed, or any other configuration. The controller 84 may be a single controller, multiple controllers , or a system 84 of controller.

Meanwhile, the controller 84 is also responsible to send command singles back to the various pumps 76, and to the auxiliary heat source 46, the auxiliary heater 70, or both. Controller 84 may control the input of heat from the auxiliary heat source 46, as well as the input of power to the compressor 50. In general, the controller 84 commands 86 or sends outputs 86 as commands 86 to the various devices and components within the system 10, and receives inputs 88 or reads 88 the inputs 88 from those and other components . In the illustrated embodiment, the controllerreceives inputs likewise from such components as the level control 48, and the level control 45. However, typically, the level controls 45, 48 operate within themselves to control the level directly, in a manner well understood in the art.

Referring to Figure 2b, a controlschema 90 identifies four levels of control. At level zero 92, the control system 90 or control schema 90 operates to control the liquid mas s . Thus , the zero level 92 may also be referred to as the liquid mass control 92. Likewise, the first level of control, above zero, is the vapor mass control 94. The liquid mas s control could operate completely independent of any othercontrolsystembut is incorporated as the basicorzero levelofcontrol in schema 90.

Likewise, the first level 94 or the vapor mas s control level 94 deals with the vapor 27 in the plenum 58, through the compressor 50, and into the closed channels 24 of the core 20. These depend on a formula relating the work done by the compres sor 50 to the pres sure and temperature within the vapor 27 pas sing through the compres sor 50. Thus , while the level zero system need only track and control a value of a liquid level, the vapor mas s control 94 has a more sophisticated responsibility. It must track the liquid levels in the liquid level controllers 45, 48, and also operate the compres sor 50 in response thereto in order to as sert control over the principle energy input to system 10, the worth of the compressor 50.

The second level control 96 or the energy control 96 is responsible for controlling a rate of change of energy inputs into the system, such as heat into the auxiliary heater 70. Accordingly, the energy control 96 must operate on the basis of a formula, algorithm, computer program, from the conditions of temperature, pres sure, concentration, and the like within the tank 30 and other components of the system 10, and as sert control over the regulation of heat through the heater 70 as part of controlling the energy of system 10.

Significant in operation of the energy control is the fact that the time of response of the tank 30 is measured in hours , sometimes many hours . By contrast, the pres sures reported by the sensors 72g, 72h in the plenum may facilitate a compres s or response in seconds . Thus, the compres sor 50 may be adjusted in current draw, and thus speed or velocity. There fo re, volumetric fiowrate can be adjusted almost instantaneously. By contrast, the addition of energy by the energy control system 96 will not be evident for a much longer period of time.

In contrast, a liquid level may be observed by sight in a manometer or gauge. However, energy flows cannot be observed physically, typically, and the rates of change and the relations hips within the system 10 are not obvious , nor intuitive.

The third level 98 of control or the system predictive control 98 is strictly algorithmic and computational in its implementation. The sophistication required is high. Many parameters , many sensors , thermodynamic considerations, material properties , and the like all go into an algorithmic determination by the system predictive control 98 of where system 10 is operating and where it should be.

Forexample, the systempredictive controlsystem 98 is responsible to review all data in the controller 84, from all sources, including the history o f o peration of the system 10. The system predictive control 98 may interpolate, extrapolate, or use other numerical method solutions to solve complex equations involving partial differentials of any value, rate of change, or the rate of change of the rate of change of variables, in order to precisely and adequately predict control set points . It may control assert control overthe heater 70, the compres sor 50, level controls 45, 48, pumps , and other volumetric flows .

The system 10 is sufficiently robust, even resilient, that it can accommodate wide variations in inputs . For example, brine concentration rates of from approximately 10,000 parts per million of total dis solved solids up to greater than 150,000 parts per millio n of total dis solved solids may be provided as inputs into the system 10. Likewise, substantially any output concentration, from such values to above 200,000 parts per million may be accommodated.

This predictive control system 98 may provide a substantial advantage to the system 10 by calculating the optimum set points for control parameters sent by way of commands 86 to the components . The system 10 may thus obtain optimum energy efficiency, brine 23 throughput to distillate 25, and so forth.

Referring to Figure 2c, a common problem in boiling regimes such as the vapor-liquid interface 100 of tank 30 between brine 23 and vapor 27 is the variable nature of the fluid level. The configuration of a meter 93 overcomes this problem. This may be important for the control schema 90.

In one embodiment of a system 10, the plenum 58 may provide a pressure source to a meter 93. The meter 93 may detect a pres sure differential, and thereby provide proces sing by the controller 84 or by imbedded proces sing, the liquid level 100 in the tank 30. Similarly, such a meter 93 may be embedded or attached as a liquid level control 45 or 48.

In the illustrated embodiment, a line 95 fromthe vaporregion, in this instance the plenum 58, will fill with vapor 27, which will condense and fill the line 95. Meanwhile, the brine 23 within the tank 30 may feed through the line 97. The two lines 95, 97 thus feed opposite sides of a gauge 93 such as manometer, of any configuration. This may be a manometer, gauge, meter, or the like. Likewise, the line 97 may serve as a common reference to other gauges 93 elsewhere in the system.

Referring to Figure 3, in one embodiment of an apparatus 10 in accordance with the invention, a tank 30 may receive input flows 35a into a diffuser 68. Those input flows 35a are received from the feed tank 32, and may pas s through a pre-treatment system 34. In one presently preferred embodiment, the brine flow 35a pas sing through the brine heat exchanger 36 receives heat from the brine 23 exiting the tank 30 through the line 33c, as controlled and driven by the pump 76d . In s uch an embodiment, the heat exchanger 36 may be set up in any one of several alternative configurations .

In one embodiment, the heat exchanger 36 may be configured as a single heat exchangerin which the flow 35a of incoming brine is counter flowing contrary to the direction of the exit brine flow 35c flowing in line 33c from the bottom of the tank 30. In such a configuration, the dwell time, heat transfer coefficient, available surface area, and the like may all be fixed, to the extent that the heat exchanger 36 may not be reconfigured.

However, in most presently contemplated embodiments , the flowrate 35a and its corresponding flow rate 35c may be used as control variables . As in Figure 2B, the control of energy typically includes the control of heat addition to incoming brine 23 preheated by the heat exchanger 36. Meanwhile, the zero level 94 from Figure 2B includes the level control. One of those level controllers 48 controls the liquid level 100 of the brine 23 in the tank 30. Accordingly, the flow 35a into the tank 30 may be used, as a control variable. As explained, the flow rate 35a through the distillate heat exchanger 44 with the corresponding output flow of distillate 25 through the distillate heat exchanger 44 may be fixed and matched to one another. The mass flow rate for adjusting the level of brine 23 in the tank 30 may be that control by the level control 48, altering the flow rate of the F2 through the pump 76b and the heat exchanger 36.

Thus , in conditions wherein the incoming flow 35a through the brine heat exchanger 36 is comparatively low F2 may reduce to less than o ne third of the flow through the distillate heat exchanger 44. In such an embodiment, relatively little heat exchange surface area is required. Thus , reduction to a single heat exchanger 36 may be appropriate.

Circumstances wherein the brine heat exchanger 36 receives a greaterproportion of flow in F2 than the distillate heat exchanger 44 receives from F l, the brine heat exchanger 36 may instead carry two or more times the volumetric flow rate of the incoming flow 35a compared to that of the heat exchanger 44 with its controlling pump 76a.

Thus , it may be advisable to provide longer dwell times , greater surface area, or both during conditions when a greater flow rate (comparatively) pas ses through the brine heat exchanger 36, than the distillate heat exchanger 44. Likewise, as flows change, the number of heat exchangers , the area available , the dwell time, or some combination thereof may be varied.

Referring again to Figure 3, multiple heat exchang ers 36 may be configured in either a series or parallel configuration, valving systems may be provided to engage one, two, three, or more heat exchangers 36 in a parallel configuration. In this way, the number of heat exchangers needed may be engaged, without subjecting the flow 35a, or the flow 35c to exces sive distance, and therefore additional fluid dynamic drag to be overcome by the power of the pumps 76b, 76d.

In contrast, flows may be slowed, and dwell times increased, while also increasing the available surface area by arranging heat exchanges in a series configuration. In a series configuration, pressure los ses may be comparatively larger. Also, valving cannot be used to direct flows between heat exchangers 36, as all flows pas s through all heat exchangers 36.

Depending upon the range of operationalparameters to which a system 10 may be subjected, a single, multiple, series , or parallel arrangement of heat exchangers 36 may be configured in the lines 33a, 33c in order to accommodate heat transfer between the flows 35a, 35c.

By way of reference, in one embodiment of an apparatus and method in accordance with the invention a sixfold variation in flow rate through the brine heat exchanger 36 neces sarily changed the flow speed and, the flow profile. As such flows may be partially laminar and partially turbulent. As will be appreciated b y those skilled in the art, such variatio n s affect the net dwell time during which heat transfer can take place, the log mean temperature difference existing between the flows 35a, 35c in the heat exchanger 36, and so forth.

Therefore, the distillate heat exchanger 44 may be designed for the flow rate output for which a system will be operated continuously. In contrast, the brine heat exchanger 36 must be tasked with the control process responsibility of matching the net flow through the systemaccording to the brine concentration ratio of incoming to outgoing brine 23.

Continuing to refer to Figure 3, specifically, while referring g enerally to Figures 1-12, a diffuser 68 in an apparatus 10 in accordance with the invention may be responsible to introduce the flow 35a into the tank 30. It has been found that several configurations may be considered, each with a somewhat different effect. Inasmuch as the tank 30 establishes a gradient of concentration from the lowest concentration of total dis solved solids at the top ofthe liquid levelin the tank 30 to a highest concentration of dis solved solids at the bottom of the tank 30. Two mechanisms tend to operate to exchange heat and mass between the incoming flow 35a and the brine 23 in the tank 30 itself.

By virtue of initial velocity of introduction ofthe flow 35a into the tank 30, momentum transfers between the incoming flow 35a and the substantially quiescent brine 23 in the tank 30. Thus, mas s may be exchanged between the jet and its consequent plume and the brine 23 in the tank 30. Momentum transfer occurs as the jet interacts with the surrounding brine 23, thus mixing, broadening, and increasing the concentration in the jet, as it mixes with the brine 23 in the tank 30.

Likewise, another mechanism, entirely different there from in its motivating force and energy, is the brine density plume. A buoyance difference between the more dense brine 23 in the tank 30 and the less dense introductory brine flow 35a from the feed tank 32 results in a buoyant force on the incoming brine flow 35a. Accordingly, the brine flow 35a tends to rise as a lighter fluid 35a within the heavier brine 23 ofthe quiescent tank 30.

This rise also results in a velocity upward by the incoming brine flow 35a, resulting in a plume with aspects of the jet-like behavior. For example, the rising, lighter flow 35a rises through the heavier quiescent brine 23 in the tank 30, mixing therewith, broadening the plume, entraining surrounding brine 23, and resulting in an exchange of momentum as well as content (dis solved solids).

A function of a diffuser 68 is to reduce the effect of a velocity -based momentum jet from the incoming velocity ofthe flow 35a. Nevertheles s, in certain embodiments , the diffuser 68 may simply be replaced by a jet.

In Figure 3, the line 33 may connect to a diffuser 68 in which the line and the diffuser 68 are both circular in cros s section. For example, the diffuser illustrated in the top embodiment illustrates an expansion ofthe diameter from the diameter ofthe line 33 as a bell, such as a bell on the trumpet.

Thus , the effective cross sectional area is gradually increased, resulting in a commensurate decrease in the velocity of the flow 35a introduced by the diffuser 68. In the illustrated embodiment, the wall 3 1 is penetrated for installation ofthe diffuser 68. Thus , the diffuser 68 introduces the flow 35a through the wall 31.

In the schematic diagram of Figure 1, the diffuser 68 is illustrated below the core 20. Each potential location has benefits .

The diffuser 68 presents no horizontal surfaces . It provides , in fact, no acces sible surfaces on which descending materials from the concentrated brine 23 ofthe tank 30 may accumulate.

The middle embodiment ofthe diffuser 68 illustrated in Figure 3 is configured more in a fan-like shape in which the net area ofthe line 33 is increased in the diffuser 68, but not with a circular cross -section. Here, the thicknes s and width ofthe fan-like diffuser 68 may be selected in order to provide a flow velocity for the flow 35a as desired.

In one embodiment, such a diffuser 68 may be oriented to discharge the flow 35a in a vertical direction below the core 20. In another embodiment, the rectangular cros s-section ofthe outlet ofthe diffuser 68 may be configured to be a square, and may c ov er a comparatively larger fraction of the area under the core. However, in the illustrated embodiment, the diffuser 68 discharges the flow 35a directly through the wall 31, and does not present any of its structure within the tank 30 itself. The lower configuration of a diffuser 68 in Figure 3 may be constructed in any of several arrangements . The illustrated embodiment shows the line 33 ported directly through the wall 31, resulting in a jet flow 35a into the tank 30. Of course, any degree of change in the cross -sectional area from the line 33 to the output of the diffuser 68 may be selected and may be appropriate. Just as the other embodiments may be arranged to pas s the flow 35a through the wall 31, or upward into the core 20 directly, from below the core 20, this embodiment may be arranged in any such manner.

In fact, the flow 35a may be directly horizontally vertically, or obliquely with respect to the bottom of the core 20. In some embodiments, the flow 35a may be introduced through a plate with apertures , through a plurality of lines 33, through various diffusers 68, through a bank of diffusers , or the like. Nevertheles s , in the illustrated embodiments of Figure 3, the diffusers 68 remain outside the wall 31. Here they are able to further reduce the components subject to the destructive forces of the concentrated brine chemistry. They also reduce the tendency toward scaling, fouling, accumulation of precipitants , and the like.

Referring to Figure 4, while continuing to refer generally to Figures 1-12, a system 10 in accordance with the invention may include channels 22 open to the surrounding tank from a top liquid level 100 in the tank 30, to a lowest outlet level 101. Meanwhile, each panel 102 around each closed channel 24 forms a mechanical barrier between the vapor 27 and condensate 25 within the closed channel 24, and the brine 23 in open channels 22, which are effectively contents of the tank 30.

Each panel 102 presents an outer surface 104 in contact with the brine 23 in the open channel 22. An inner surface 106 of the wall 26 is in contact with the vapor 27 or condensate 25 in the inner or closed channel 24. Heat is transferred from the high pres sure region, having a higher saturation pres sure and higher saturation temperature in the closed channel 24. As described with respect to Figure l, the compres sor 50 compres ses the vapor 27 from the plenum 58 to a higher pressure, and corresponding temperature in accordance with the Clausius -Clapeyron equation illustrated in Figure 7C. The lower pressure and temperature region is represented by the plenum 58 and the open channels 22 in the tank 30.

Thus , heat transferred from the closed channel 24 p as s e s through the wall 26 subject to the heat transfer coefficients on the inner surface 106 and outer surface 104 of the wall 26. Ultimately, convection cells due to thermal convention may occur. Thermal convection is the result of buoyancy, a density decrease by a fluid that has been heated compared to its surrounding and comparatively cooler neighbors .

For example, to the extent that the tank 30 represents brine 23 that has been stratified, stratification in response to b rine density differences is several times more significant than the buoyance differential due to a temp erature difference. Accordingly, hotter brine 23 may still remain lower, or at a lower level, within the tank, due to the fact that its dissolved solids content prohibits its rising in response to thermal buoyance effects .

Nevertheless , buoyance differentials due to heat addition, rendering the hotter material to be o flower density, and thus lighter, may result in Benard cells 108 or Benard convection cells 108. However, these will exist only locally within material having the same density with respect to dis solved solids . The cells 108 tend to move heat from the wall 26 into the bulk of the brine 23 in the open channel 22.

One advantage of a profile (e.g., brine concentration gradient or a density variation with depth) due to dissolved solids concentrations increases with depth within the tank 30 and the open channels 22 is the fact that rather than rising immediately alo ng the wall 26, heated brine 23 may remain localized, thus contributing to increase temperature at a comparatively lower level.

In this way, heat may be transferred continually from the wall 26 into the brine 23 of the open channel 22, even though the temperature differential between the vapor 27 in the closed channel 24 may be nearer to the temperature of the adjacent brine 23 in the open channel 22. Heat transfer still continues because the convection cells 108 did not neces sarily become general along the entire height 1 16 of the panel 102. Rather, energy is "pumped" away from wall 104.

Stated another way, brine 23 at a particular level in the open channel22 may stillcontinue to pick up heat, and may cause a generation of bubbles 100 at the outer surface 104 of the wall 26, which might otherwise not be able to occur. Compared to the illustrated embodiment and the apparatus in accordance with the invention in free convection, if the tank 30 were full of clean water, heated liquid would always rise in the presence of comparatively cooler liquid. Thus , all the hottest liquid would rise to the top .

In contrast, with stratified brine, hot liquid may exist and remain at the bottom. In fact, a reverse temperature gradient, in which the hottest temperature is at the lower end of the panel 102 is entirely pos sible, depending on the heat transfer dynamics of the system 10.

In general, bubbles 1 10 are generated at the outer surface 104 of the wall 26 of the panel 102 anytime localized brine 23 achieves the saturation temperature for its localized pressure. Pres sure varies with depth, and density of the brine as well as the overhead pres sure within the plenum 58. Thus , lower in the open channel 22, one expects and observes higher pres sure.

Moreover, due to the density pro file (e.g., gradient) orconcentration pro file (e.g., gradient), saturation pres sures and temperatures rise even further.Nevertheless , inasmuch as the temperature within the closed channel 24 is higher than the temperature in the open channel 22, heat transfer may still occur acros s wall 26, and bubbles may be generated at the lower extremities of the panels 102.

This phenomenon has been observed in practice during experiments . For example, in free convection with a condensing vapor 27 within a closed channel 24, wherein the outerchannel 22 or open channel 22 contained no saline gradient, the formation ofbubbles 100 occurred only within the top 5 percent ofthe height 1 16 of the panel. In contrast, b ub b le fo rmation was observed within the bottom 20 percent of the open channel 22, when the open channel 22 contained stratified brine 23.

As each bubble 100 is formed, it would typically nucleate at a site on the outer surface 104 of the wall 26 of the panel 102. However, it will quickly separate as it grows, and move from a position illustrated by the bubble 100a to a position in the free stream of the open channel 22 illustrated by the bubble 100b.

It has been observed that as bubbles 100 grow, due to heat addition, mass addition, and even due to a simple rise in altitude reflecting a reduced surrounding pres sure,thebubbles lOOhavebeen observed to strip the boundary layer fromthe surface 104 ofthe panel 102. This triggers the generation of clouds ofbubbles as illustrated by the bubbles lOOd of Figure 4. These bubbles lOOd likewise appear to be able to grow and rise. Nevertheless , they may not neces sarily nucleate at the wall 26 but may be generated b y an infusion of heat due to the disruption of the thermal and fluid boundary layer as understood in the art of heat transfer. As the bubbles 100 continue to rise, they tend to grow in size, and tend to coalesce with one another. They begin to form larger bubbles , and tend to move toward the brine 23 in the open channel 22, and away from the wall 26. In fact, as a practical matter as a bubble flow 112 rises , brine is displaced , and a corresponding downward flow 114 of the surrounding brine occurs . A simple mas s or volumetric analysis illustrates that as mas s rises in the open channel 22, a certain amount of the mas s must go down and take its place. This results in a flow 114 around each bubble 100, as illustrated.

As a result of the formation of each bubble 100, vapor 27 leaves the brine 23. Salt, the chemicals listed hereinabove that may be contained in the brine, and the like, may be volatile and nonvolatile. Such contaminants as methanol, may evaporate into the vapor 27. However, salts , dissolved solids , and the like must remain behind and do not evaporate.

Accordingly, the flow 114 around each bubble 100, at the time of formation of the bubble 100, neces sarily receives the dissolved solids that cannot vaporize. Experiment s on apparatus and methods in accordance with the invention demonstrate a downward flow 114 ofheavierbrine, resulting in a net gradient having the lowest concentration of dis solved solids at the top surface 100 of the liquid, or the liquid level 100 and the highest concentration of dis solved solids at the bottom of the tank 30.

In general, the height 116 of the panel 102 may be selected to optimize heat transfer. Likewise, the distance or thicknes s 118 acros s the wall 26 may be selected for structural and thermal considerations . Similarly, the width 120 of the open channels 22 may be selected in order that the bubbles 100c coalescing together do not obstruct the channel 22, nor dry the outer surface 104 of the panel 102. Such drying of the surface 104 may result in additional scaling, and has been observed to exacerbate corrosion of the wall 26.

The width 122 or thickness 122 of the closed channel 124 may be selected to optimize heat transfer and permit flow by naturalconvection, thus limiting or euminating the need of conventional heat exchange, wherein pump energy is used to drive all flows . By contrast, in the illustrated embodiment, the open channel 22 operates by a saline convection or dis solved solids convection with thebrine. This is based on buoyance differentials between various flows and regions of the brine 23. Similarly, the vapor 27 within the closed channel 24 as it condenses on the inner surface 106 of the panel 102 eventually forms a condensate 25 collecting at the bottom thereof and exiting out the plenum 19b for liquids .

It has been found that the height 124 or distance 124 between the top of the panel 102 and the liquid level 100 may be positive. In some embodiments , it has been found that heat transfer rates may be effected by vigorous boiling of bubbles 100c near the top of the p anel 102. It has been found most effective in the presently contemplated embodiments , as demonstrated by experiments , to maintain the liquid level 100 above the top of the panel 102.

The c o re 20 will typically be spaced a distance 126 from the outlet level 102 of the tank 30. Typically , a significant volume in the plenum 58 above the liquid level 100 tends to provide a volume against which the compres sor 50 may draw. Similarly, a larger depth 126 between the tank outlet level 101, than herein illustrated schematically is desired.

The height 126 is illustrated by a cut line indicating that any additional distance may be added therein. Though not shown in the illustration, such an addition provides the pos sibility of increasing of highest density brines 23 from the open channel 22 toward the bottom of outlet level 101 of the tank 30. Nevertheless , the activity within the tank 30, and specifically when the open channels 22 is a densification or increase in concentration of dis solved solids in the brine 23. It has been found generally that the region, illustrated by the height 116 of activity of the concentrating proces s , is the region that sees the largest change in density profile. Accordingly, the density within the height 126 of the region below the panels 102 does not show the intensity of the steepnes s of gradient.

Inasmuch as the plenums 19a, 19b carry differentialdensities , they are different sizes . In fact, the upper plenum 19a may be thought of as simply a manifold 19a feeding vapor at a comparatively larger specific volume, lower specific density, into the closed channel 24. Similarly, the condensate 25 has a density almost 1,000 times greater than that of the vapor 27, corresponding to a specific volume of about one thousandth of the volume of the vapor 27. Thus, the manifold 19b or plenum 19b receiving condensate 25 from the closed panel 27 need not have the same volumetric capacity as the upper manifold 19a.

In general, droplets 130 form against the inside surface 106 of the wall 26 in the panel 102. Droplets 130 tend to migrate downward and may likewise coalesce into streams orrivulets running into the condensate 25 collected at the bottom of the panel 102, resulting in condensate level 128 accumulating in panel 102. The result of condensing vapor 27 on the inner surface 106 of the wall 26 of the closed channel 24, is a very high heat transfer rate on the order of 20 times greater than the heat transfer rate between liquids across a solid surface.

Thus , for example, the rate of heat transfer into the brine 22 from the outer surface 104 ofthe wall 26 is lower when merely resulting in heat transfer into the liquid brine 22. In contrast, the heat trans ferrate, and thus the heat transfer coefficient upon nucleate boiling with bubble 100 formation is comparatively about 20 times that rate, and corresponds to the condensation heat transfer rate on the inner surface 106 of the wall 26 in the panel 102 enclosing the closed channel 24.

Referring to Figure 5, a chart 134 illustrates axes 136, 138. The height axis 136 illustrates the height from the outlet level 101 ofthe tank to above the liquid level 100 including the plenum 58. Meanwhile, the TDS axis or the total dis solved solids axis 138 illustrates the concentration oftotal dis solved solids within the tank.

The curves 140 are gradients or pro files of concentration or density. In the illustration of Figure 5, the location of the core 20 is shown in dotted lines, as is the outer shape of a tank 30. The outer level of the tank 30 is illustrated, along with that ofthe core 20 in order to show the response ofthe density profile 140 to the altitude or height 136 along the height axis 136.

In the chart 134 o f Figure 5, the curves 140 represent concentrations varying from a minimum amount corresponding to the value on the TDS axis 138 at its minimum value, at its intersection with the vertical axis 136 or height axis 136. Meanwhile, at the liquid level 100, the concentration and therefore the density of brine 23 in a tank 20 is at a minimum value of dis solved solids in an apparatus and method in accordance with the invention.

However, in a mixed environment, one in which the brine 23 in a tank 30 is completely mixed, a profile 140 reduces to a vertical line, having a constant concentration and constant density throughout fromthe liquid level 100 to the outlet level 101. Thus , the concentration at the liquid level 100 is the same as that at the outlet concentration 142. In an environment in which a gradient profile may be established ideally, a static linear density profile may be established according to the curve 140b . In this situation, the concentration varies from a minimum value at the liquid level 100 and increases to a maximum outlet concentration 142 at the outlet level 102.

In order to establish the ideal gradient illustrated by the profile 140b, it would be neces sary to maintain a continuous, and equal change in concentration at substantially every level between the outlet level 101 and the liquid level 100. This would require tremendous control, though it would also provide a predictable and useful and consistent gradient in the tank 30.

Experimental results in an actual apparatus 10 in accordance with the invention is illustrated in the dynamic density gradient profile 140c. In this profile 140c is seen the change in the density gradient within the core region, as compared with the change indicated in the region below the core. Between the liquid level, the bottom of the core 20, and the outlet level l01, the normalized concentration difference, from the lowest concentration level at the liquid level continually increases to the outlet level 101.

Ac c o rding ly , the outlet concentration level 142 of both profiles 140b and 140c originate and terminate at equivalent points . In contrast, however, the density gradient curve 140c is shown to stabilize in a different shape, in which mos t o f the c o nc entration increase occurs within the altitude of the core 20, and very little change occurs therebelow. Thus, the region of the tank 30 below the core 20 may still maintain a gradient.

However, in these experiments not nearly so substantial a total difference as that achieved within the core 20 was observed. This is seen as indicating severalfacts, including the fact that the core 20 is the region in which the open channels 22 are concentrating brine 23 by evaporating off vapor 27. Below the core, where no substantial vaporizing occurs , the difference in concentration is substantially les s .

In reviewing the chart 134 of Figure 5 in view of the phenomena illustrated in Figure 4, one may ascertain the degree of mixing occurring within the open channels 22, as opposed to elsewhere in the tank 30. Also, to the extent that the diffuser 68 of Figures 1-3 is located below the core 20, brine convection, or the brine density buoyant convection, will occur.

Likewise, for example, brine 23 incoming in the input flow 35a is lighter than any brine 23 within the tank 30. Thus , regardless of the velocity with which the flow 35a is introduced into the tank 30, it will immediately begin to rise through the core 20, or anywhere else within the tank 30 that it is introduced.

Accordingly, the brine buoyance plume created by the inlet brine flow 35a will rise toward the liquid level 100, exchanging momentum, mas s density, and heat with the surrounding brine 23 through which it pas ses . The dynamic density gradient profile 140c therefore illustrates that the actualvalue of concentration or density within the tank 30 is neither the ideal static linear density profile 140b, nor is it the mixed non-gradient 140a.

Thus , the dynamic density profile 140c (gradient 140c) is very useful in the control and stabilization of a system 10 in accordance with the invention. Of course, the ideal static linear density gradient 140b would be very useful but difficult to achieve, maintain, orboth. However, experiments at this point demonstrate that moving away from the mixed, non-gradient condition illustrated in the curve 140a can be achieved, are easily maintained, and provide very useful results . Referring to Figure 6, a chart 145 illustrates curves 146 of an increase in total dis solved solids (TDS) as a function of the feed concentration thereo f. The formula 148 illustrates a normalized total dis solved solids increase represented by each of the curves 146. The curve 146a represents the increase in total dis solved solids in the brine 23 in the tank 30 at the comparatively minimum feed concentration of dis solved solids in experiments with the apparatus 10 in accordance with the invention.

In contrast, the curve 146e illustrates the increase in the normalized total dis solved solids content in the brine 23 of the tank 30 at the comparatively highest input concentration of dis solved solids in the experiments . The height axis 136, as in Figure 5, again measures from the outlet level 101 in the tank 30, or of the tank 30 up to above the liquid level 100.

The liquid level 100 is the maximum height at which a general quantity of liquid brine 23 exists in the tank 30. Accordingly, only the plenum 58, holding vapor 27, exists immediately above the liquid level 100. Thus, the dis solved solids content has a value 150 at the liquid level 100. In the experimental system 10 in accordance with the invention, the lowest concentration of dis solved solids occurs at the liquid level 100. Thus , allvalues measured along the axis 138 are normalized against that minimum concentration of dis solved solids 150.

The shape ofthe curves 146 reflects the change in rate of concentration increase with depth toward the outlet level 101. Thus, curves 146b, 146c, 146d reflect intermediate input TDS curves within the family of curves 146. The experimental data is contained in curves 146a, 146e. However, the c o n s istent curvature obtained through multiple experiments illustrates that the concentration pro file and gradient within the tank 30 are independent from the output TDS.

For example, the curve 146a corresponds to input feed concentrations of 50,000 parts permillion as well as feed concentrations of 100,000 parts per million (ppm). Likewise, the curve 146e represents tank concentrations of 100,000 ppm and 200,000 ppm output brine concentrations . However, the input dis solved solids concentration 146, when closer to the outlet concentration of dis solved solids , appears to have les s effect on mixing.

Likewise, the larger the discrepancy between the concentration at the inlet flow 35a compared to the outlet brine flow 35c shows a tendency ofthe more concentrated brine in a tank 30 to rapidly dampen the effect on concentration by the incoming flow 35a. Thus , as the input TDS increases , the curve 146 moves fromthe curve 146a toward the curve 146e.

Meanwhile, minimum and maximum output concentrations of dis solved solids both result in the curve 146a. Thus , the normalized TDS increase is independent from the output concentration of total dis solved solids at the outlet concentration 152 at the outlet level 101 in the tank 30.

The curve 146a corresponds to four sets of experimental data. Two experiments involved vaporrecompres sion in an apparatus in accordance with the invention at a 50,000 ppm of brine input into a tank 30. In o ne p air of experiments , the output brine concentration was near orat 200,000 parts per million, the other 100,000 parts per million. Meanwhile, the curve 146e corresponds to two experiments in which the output TDS concentration was 200,000 parts per million. The input feed rate was 100,000 parts permillion in each ofthose experiments, and the outputs were 180,000 parts per million and 200,000 parts per million, respectively. Referring to Figure 7A, a chart 154 illustrates a distribution oftemperature measured along the axis 156 against a height measured along the axis 136. In the chart 154, the average tank temperature 158 is illustrated at various positions, including the value of T l or first temperature identified by the line 158a, and a second temperature or T2 at the line 158b. Here, the curve 160 reflects the saturation temperature in a stratified concentration profile of the tank 30.

The curve 162 illustrates the saturation temperature of a mixed brine 23 in a tank 30. The difference between these curves 160 and 162 reflects the difference in saturation temperature as a function of stratified brine concentration versus completely mixed brine in accordance with Raoult's Law, illustrated in Figure 182. Both illustrate changes in saturation pres sure with depth and density. Therefore, the curves 160, 162 accommodate the depth difference at various locations within the tank 30.

Reference to Figures 7A-7E are best understood when viewed together. Figure 7A is a chart 154 illustrating saturation temperature differences between a saturation temperature in a stratified tank (curve 160), which is not a line, but rather a non-linear curve; the saturation temperature curve 162 describes a completely mixed tank 30. Thus , these two curves 160, 162 correspond to the dynamic density profile curve 140c of Figure 5 and the mixed non-gradient curve 140a of Figure 5, respectively.

The difference between the two curves 160, 162 is best understood by reference to Raoult's Law 182 described in Figure 7B. Here the equation states that the change in saturation temperature within a brine is equal to the product of the ionic constant corresponding to the chemical constituents making up the brine, with the ebullioscopic constant corresponding to the units of degrees Celsius times kilograms divided by moles for water.

Similarly, the Clausius -Clapeyron equation 184 de s cribes the rate of change of pres sure with temperature according to the dependents on the latent heat of vaporization divided by the temperature and the change in specific volume (volume per unit mas s). This equation may be written in several forms including one that indicates the change in pres sure is equal to the pressure within the bulk fluid times a power of the natural log. In this last embodiment, the coefficient 'm' is at best isolated as the lower version of the equation 184 in Figure 7C.

Thus , Raoult's Law governs the change in saturation temperature due to the impurities within a liquid. The Clausius -Clapeyron equation corresponds to the change in pres sure as a function oftemperature due to compression of a vapor.

Figure 7D illustrated Dalton's Law 186, sometimes referred to Dalton's Law of Partial Pres sures 186. Here, pres sure within any volume is equal to the fraction ofthat volume occupied by any particular vapor, usually an idealized gas , multiplied by the vapor pressure of that gas . Thus , the pres sure in the plenum 58 is a combination ofthe partial pres sures of all the evaporated vapors 27 existing therein.

Referring to Figure 7E Henry 's Law 188 describes the relationship between concentration of a solute (dis solved gas) dis solved in a solvent. Accordingly, the concentration of a particularly species designated by the lower case letter 'i' is a function ofthe partial pres sure or vapor pres sure of that constituent in a volume divided by the Henry's Law constant.

Thus , Henry 's Law 188 describes how much of a supposedly non-condensable gas is actually absorbed. Henry 's Law also applies to other condensable gas ses . Therefore, when considering Figure 7A, the s turation temperature within the tank 30 corresponds to a saturation pres sure at any point (depth) within the tank 30. However, the saturation pres sure varies with the constituents dis solved in the brine, the volatile ions thereof, and the depth at which one is observing temperature, pres sure, and so forth, in accordance with the foregoing equations .

Still referring to Figure 7A, while continuing to refer generally to Figures 1-12, the T l and T2 average tank temperatures 158a and 158b merely s erv e as p o ints of reference. The effects of the curves 160, 162 apply at any particular location or depth within the tank 30. Accordingly, the tank temperature 158 is significant to a local effect on the saturation temperature required to boil liquid to vapor 27 in the channel 22. Thus, the significance of the temperatures 158a, 158b is actually their relationship to the localized saturation temperature as given by curves 160 and 162.

Effectively, the curve 162 is a calculated value corresponding to the saturation temperature of the brine 23 in a fully mixed tank 30. Thus, the brine 23 corresponding to the curve 162 is fully mixed and corresponds to the profile 140a as illustrated in Figure 5 and illustrates the absence of a profile or gradient.

At a tank temperature 158a, the saturation temperature 162 corresponds to boiling liquid level 100 or a boiling surface point 164. Remaining in a mixed condition and descending from the liquid level 100 down toward the outlet level 101, the saturation temperature 162 rises to a maximum of 156. This rise is due to the head level or height of the liquid column above any particular location along the curve 162.

In this example, the surface boiling point 164 is at the surface exactly because there is no submersion below the liquid level 100, so the saturation temperature 162, by the definition of saturation temperature, occurs at the boiling surface 164. In this example, the saturation temperature curve 162 takes into account Raoult's Law 182 and its effect on the boiling temperature 162.

If the tank temperature is raised from the value 158a to a higher temperature 158b, then the same head height is imposed by the liquid level 100. Thus, the new boiling point 166 corresponds to surface boiling into the plenum 58, by the brine at the temperature value 158b . Accordingly, if the temperature profile 162 or temperature curve 162 were shifted to the right, corresponding to the increased temperature 158b, the curve 162 would only go through the point 166 if the saturation pressure in the plenum 58 also rose to the appropriate saturation pres sure.

If the saturation pres sure in the plenum 58 does not rise, then the region between the liquid level 100 corresponding to point 166, and the height along the axis 136 corresponding to the point 168 willallbe boiling. In other words , core nucleate boiling would occur in the upper core region of 176.

Nevertheless , in another example, one may think of the tank 30 being at the average temperature 158b and having the compres sor 50 draw down the vapor 27 in the plenum 58. This would result in the drop of the saturation pres sure. Accordingly, having the plenum 58 at the saturation pres sure corresponding the curve 162, while the average tank temperature is at 158b, the core region 176, between the surface point 166 and the point 168 on the curve 162, boils generally throughout.

The curve 160 represents the saturation temperature existing in the a dynamic density profile orgradient 140c in a tank 30. The curve 160 intersect s the tank av erage temperature 158a at a point 170, a condition wherein the saturation temperature 160 within the tank 30 is exactly at the tank average temperature. Likewise, the point 172 corresponds to the intersection ofthe saturation temperature curve 160 and the elevated tank average temperature 158b . These two tank temperatures 158a, 158b are illustrated as constant throughout the altitude ofa tank 30 and are used merely by way of example. With a dynamic density profile 140c (see Figure 5) the stratification within the tank 30 may create any one of a variety oftemperature profiles . It is also possible to have a non-constant temperature throughout the height of the tank 30. In other embodiments it may be pos sible to have reverse gradient s in which the hottest temperature is at the bottom of the tank.

Again, anytime the localized saturation temperature 160, or 162 is below a localized temperature 158a, 158b or the like, the brine 23 in the tank at that location will be in boiling mode.

In Figure 7A, assume that the pres sure in the plenum 58 is always at the same, constant value throughout the following discus sion. In the chart 154, one may select a temperature 158a within the tank 30.

Now, cons iderthe saturation temperature curve 160 occurring with a gradient in accordance with the invention and represent s the dynamic density profile 140c of Figure 5 due to stratification of the brine 23 in the tank 30. The intersection of this curve 160 intersects the surface 100 not at the point 164, but at some point lower in temperature on axis 156. In accordance with the foregoing discussion, boiling now begins as lo w as the point 170, where the tank temperature 158a intersects the saturation pres sure 160 of the brine gradient 140c.

Given this condition existing at point 170, the entire region 174 above point 170 is in a full boiling condition, with no additional energy introduced. In the configuration represented by the chart 154, the fully mixed saturation temperature 162 is set to boil at the surface 100. In contrast, at the same tanktemperature 158a, the stratified brine boils in the entire region above the point 170. Thus , more ofthe core is involved in high heat transfer nucleate boiling, as compared with that which would have been achieved in a completely mixed system.

Of course, if temperature were raised in order to engage more of the core 20 in boiling, as would correspond to increasing the tank temperature to the value 158b, the point 166 is the temperature value 158b required in the tank.

However, in the stratified condition corresponding to the curve 160, the point 172 reflects the point above which full nucleate boiling occurs throughout the core 20 in the tank 30. Thus , the region 180 represents the additional benefit, or the additional region ofthe core 20 in which full nucleate boiling is generalized in the core 20 as given by region 178.

Just as the region 174 above the point 170 represents the portion ofthe core 20 in full nucleate boiling when the tankaverage temperature corresponds to the value 158a, this increased benefit continues at allpoints along the curve 160. Regardles s of the depth of the region 176 may be, in a fully mixed tank corresponding the saturation temperature curve 162, the stratified saturation temperature curve 160 always provides an improved performance represented by region 174, 180, or by other corresponding differences between the two curves 160, 162.

This benefit may be realized in one of several alternative ways , such as the ability to run the system 10 at a reduced temperature for the same performance. Alternatively, the work done by the compres sor 50 may be reduced due to the decreased demands on the saturation pres sure in the plenum 58 above the liquid level 100.

Referring to Figure 8, an experimental system 10 was configured with a housing 12 having a motor generator system. Ultimately, an auxiliary heater relying on line power was also included. The tank 30 was set up with a core 20 placed therein connected to an upper, vapor manifold 19a, and a lower, liquid condensate manifold 19b . The core 20 included open channels 26 in liquid communication with the surrounding region ofthe tank 30, while the closed channels 24 were sealed away from the tank brine 23. A plenum 58 above the core 20 accumulated vapors boiling from the open channels 22 of the core 20. A mist eliminator 54 (not seen in Figure 8) was positioned within the plenum 58. Meanwhile, a heat exchanger 15 was installed, but was not used in the experiments reported in Figures 3-7E. The conduits 18 conducted vapor from the plenum 58 to the compres s or 50, which then passed those vapors at an increased pres sure into the vapor plenum 19a or manifold 19a.

The manifold 19a distributed the vapors 27 into the closed channels 24 of the panels 102 for condensation. Condensate 25 exited the closed channels 24 through the bottom manifold 19b as condensate. Ultimately, after holding in a reservoir 43 the distillate or condensate 25 fromthe closed channels 24 was eventually pas sed on to a distillate tank.

Above the core 20, a plenum 58 was arranged, and contained a mist eliminator 54. After pas sing through the mist eliminator, the vapor 27 in the plenum 58 was pas sed by a heat exchanger 15 which was not active during the experiments reported in Figures 4-7E. Instead, the vapor 27 pas sed on into the conduit 18 toward the compres sor 50. The compres sor 50 increased the pres s ure , and the temperature in the vapor, pas sing the vapor at this increased temperature and pres sure of saturation back into the manifold 19a at the top of the core 20.

The manifold 19a pas sed the vapors into the closed channels 24 of the panels 102, where it was condensed by discharging the latent heat of vaporization into the surrounding brine 23 in the open channels 22 of the tank 30.

The condensate 25 was then pas sed fromthe closed channels 24 into the manifold 19b at the bottom of the core 20, and ultimately discharged through a reservoir 43 into a heat recovery system 47 (as illustrated in Figure 1) to a distillate tank 42, which are not shown in Figure 8.

Meanwhile, the system was instrumented with sensors 72 near the central geography of the core 20. Heaters were positioned along the walls 31 ofthe tank 30. The motor 17 driving the compres sor 50 was controlled through a control system that would vary current to the motor 17, thus altering the velocity, throughput, and volumetric flow rate ofthe compres sor 50.

The sensors 72 were placed in the open channel 22 at the center of the core 20. Likewise, sensors 72 were placed along the walls as illustrated, and distributed schematically in Figure 1. Sensors 72 were configured to detect temperature and concentration within the brine 23 of the tank 20, in the core, and near the wall 31. Other temperatures and pres sures were detected in the plenum 58, the conduits 18 on the upstream side and downside stream side ofthe compres sor 50, and so forth. Various experiments were run on the apparatus 10 of Figure 8 in the development ofthe data of Figure 6.

Referring to Figure 9, a mas s balance reflects the input ofthe flow rates 35a introduced into the tank 20 and the output flows 35c exiting the brine heat exchanger 36, as well as the quantities of distillate 25 or condensate 25 pas sed through the distillate heat exchanger 44 to the distillate tank 42.

The experiments contributing to the charts 134, 145 of Figures 5 and 6 respectively correspond to the data obtained at the experimental conditions 194a and 194b, the conditions at location 194c and 194d, and the conditions at 194e, 194f. The conditions 194a, 194b correspond to an input feed of 50,000 parts per million. The output corresponds to a brine concentration of 100,000 parts per million in total dis solved solids exiting the tank 30.

Meanwhile, the conditions 194c and 194d correspond to an input brine concentration of 50,000 parts permillion and an output concentration of 200,000 parts permillion. Likewise, the conditions 194e and 194f correspond to an input concentration of 100,000 parts per million with an output concentration of 180,000 parts per million and 200,000 parts per million, respectively.

The data conditions of the table 190 of Figure 9 correspond to a constant output of distillate 25 of 100 barrels (168 liters) per day. Notwithstanding the output brine concentration forthe conditions 194e and 194f were not only set at 200,000, the conditions under experiment 194e were set at an output brine concentration of 180,000parts permillion.

Referring to Figure 10, the experiments corresponding to the conditions 194 of Figure 9 were implemented in the system 10 of Figure 8. The chart of Figure 10 plots the normalized increase in total dissolved solids , according to the formula 148 illustrated thereon through the six experiments 194 or the six sets of experimental conditions 194.

The normalized concentration of dissolved solids is illustrated on the TDS axis 138 and plotted against the height from the outlet level 101 up to the liquid level 100 in the tank 30. The region above the liquid line 100 or liquid level 100 corresponds to the location just above the core 20, which was completely immersed in brine 23.

The total dissolved solids value 150 at the top of the brine 23 or the liquid level 100 is normalized, or used as the normalization value, for all the flows . Accordingly, the normalized increase in total dissolved solids is expres sed as a fraction above the concentration value at the liquid level 100.

As can be seen from the chart of Figure 10, the curves 195, 196, 197 reflect the fit of data obtained. The curve 195 corresponds to the fit of data to experiment 194a and experiment 194b. The curve 196 is a fit to the experiment based on the conditions 194c and 194d. Likewise, the curve 197 is fit to the data corresponding to the conditions 194e and 194f.

The curves 195, 196, 197 correspond to the curves 186 of Figure 6. Meanwhile, F135a and F235a are illustrated by their relative height along the height axis 136. One may note that the brine buoyancy plume effect was significant in altering the total dis solved solids within the gradient in the tank 30. Additional information is also available from the raw data charts corresponding to the experiments 194 in Figure 10.

For example, at the location where the feed flows 35a were introduced into the experimental tank 30, no diffuser 68 was available. Accordingly, the flows 35 were introduced as severalpipes each injecting a horizontal jet of the input feed brine 35a from the feed tank 32. Both momentum in the horizontal direction, and the buoyancy forces vertically affected the integration of the input flow 35a into the brine 23 of the tank 30.

Moreover, the experiment conditions 194c and 194d were intended to correspond to an output b rine concentration of 200,000 parts per million. In contrast, the conditions 194a, 194b were intended to correspond to an output brine concentration of 100,000 parts per million. The effect of brine concentration in the inp ut flo w 35a is significant. Where the brine concentration in the tank 30 was four times that of the incoming brine flow 35a, the tank brine 23 very quickly rectified the concentration of the incoming flow 35a. Above the location of the feeds 35a, the curves 195, 196, 197 match quite closely the raw data.

However, in the vicinity of the incoming flows 35a, the disruptive effect of mixing is seen in the reduction of concentration below the curves 195, 196, 197. This suggests that the system 10 is very robust. For example, the curves 195, 196, 197 are highly dependent on the incoming concentration of the incoming flow 35a and exhibit virtually no dependence on the output concentration. Thus , the dynamic density profile 140c (see Figure 5) as detailed by the curves 146 (see Figure 6) may be relied upon to provide a stable, predictable output condition for the system 10. The heat input, and work into the compres sor 50, may be adjusted to accommodate the incoming feed flow 35a to reach the output desire. Significantly, the output result is not in substantial question.

The data of Figure 10 also substantiate the robust performance and resilience of the gradient in the tank 30 in the face o f wide variations in the incoming brine concentration. This is particularly significant in actual production facilities where the incoming brine flow 35a may v ary as fracture water, production brine, or the like. These data demonstrate that the output and control of the system 10 need not be subject to such arbitrary inputs .

Referring to Figure 11, a chart 198 illustrates the effect on temperature along the temperature axis 150 at various levels of depth illustrated on the axis 136. Temperatures are not normalized to a non-dimensional form as with other figures . Here, the saturation temperature curves 160, 162 correspond to those of Figure 7A . These values reflect actual data from the experiments 194 corresponding to Figures 9-10. Here, the point 199 represents the saturation temperature at the surface 100 or the liquid level 100 in the tank 30. The co nditio n s at the point 199 constitute the saturation temperature at the liquid level 100 for a fully mixed tank. This corresponds to the conditions of the curve 162.

Similarly, the point 200 represents a saturation temperature at the pres sure in the plenum 58 above the liquid level 100. Likewise, saturation conditions at the liquid level 100 along the curve 162, the point 200 corresponds to the curve 160 of a saturation temperature existing at the pressure in the plenum 58 for a dynamic gradient configuration of Figures 9-10.

The empirical data of Figure 11 confirm the operational characteristics discus sed with respect to Figure 7A. For example, the region 176 corresponds to the description of the region 176 with respect to Figure 7A. Likewise, the region 178. Similarly, the region 180 of Figure 7A is the difference between the depth of the region 176 and the region 178.

This represents the advantage in heat transfer area and greatly multiplied heat transfer coefficient in the region of nucleate boiling in the core. The core 20 is illustrated by dotted line surrounding a region reflecting the actual depth and position of the core 20 in the tank 30 during the experiments 194.

Referring to Figure 12, while continuing to refer generally to Figures 1-12, a proces s 202 for controlling an apparatus 10 in accordance with the invention may have several levels of control. For example, a zero level 203 represents balancing mas s by a continuous proces s of tracking and adjusting the levels of liquid in the tank 30 and in the distillate reservoir 48. These are directly observable and adjustable timely by conventional measurement and control techniques .

Meanwhile, a level one co ntrol 204 as well as a level two control 205 and a level three control 206 are illustrated. Level one control 204 involves controlofthe work done by the compress or 50. In the illustrated proces s 202, level one is seen as intervening 204 in the operation of the proces s 202 operating in the system 10.

A principalmechanismforcontrolis reducing 211 orotherwise changing 211 the work done by the compressor 50 in removing vapor 27 from the plenum 58. Typically, the reducing 211 operation corresponds to a control intervention 204 initiated in response to an undesired rise in the liquid level of the reservoir 43 containing distillate. A change 211 in the work done by the compressor 50 causes a response 212 in the system 10. For example, reducing 211 the work done by the compres sor 50 causes arising pressure in the plenum 58. Likewise, a decreasing mas s flow rate willresult through the compressor and out of the plenum 58. This is somewhat counterintuitive.

For example, decreasing 211 or reducing 211 the work done by the compres s or backs up pres sure in the plenum 58, causing core boiling to decrease and decreasing the distillate temperature and saturation pressure. These system responses 212 result in a readjustment of the operation point of the system 10. Specifically this alters the pres sure in the plenum 58, thereby forcing a readjustment of saturation pres sure and saturation temperature in the brine 23.

As a practical matter, the proces s 202 illustrated in Figure 12 is an examp le of controlling the system 10. Accordingly, the most responsive element for controlling operation of the apparatus 10 or system 10 is the level one control 204.

Level two control 205, or intervening 205 in the level two control scheme, involves adjusting 213 the auxiliary heat provided by the auxiliary heater 70. In this example, adjusting 213 is embodied in decreasing auxiliary heat output by the auxiliary heater 70. This may be done by controlling the heater 70 or the auxiliary heat source 46.

By decreasing 213 auxiliary heat, intervening 205 follows the more rapid and responsive 211 ofthe intervention 204. However, in intervening 205 at level two control, the decrease 213 in auxiliary heat result s in a much s lower response ofthe system. This includes a decreasing temperature in the tank, decreasing mas s flow rate ofthe distillate 25, and decreasing core boiling.

The temperature 158a is moved to the left in Figure 11. However, an excursion away from the curve 162 will typically occur as a system response 212. By decreasing 213 the auxiliary heat, the temperature line 158a moves to the left, corresponding to a net cooling of the tank 30. The result of moving the line 158a to the left is a decrease in the region 176 and an increase ofthe region 180 between the regions 176, 178.

Perhaps the most significant effect of moving the temperature or decreasing 213 the heat with its corresponding decrease in the temperature 158a is to reduce the region 178, by shifting the position ofthe intersection point 172 at which the temperature line 158a intersects the curve 160. Thus , les s ofthe core 20 is involved in boiling. Accordingly, a decreasing mass flow rate and a decreasing core boiling willoccuras systemresponses 214 in accordance with Figure 12.

Intervening 206 at the level three control, as described in reference to Figure 2B, may involve proces sing 215 by a computer pro ces s or in order to provide a predictive trim to the other levels of control. Accordingly, the controller 84 may receive signals from any or all ofthe sensors 72. It may provide instructions commanding 216 modifications to the work, heat, or, optionally, mass flow rates in the system.

Commanding 216 an alteration to these independent control variables may result in alteration ofthe dependent variables . Accordingly, feeding 217 data back orproviding 217 feedback of values of pres sure, temperature, mas s flow rate, concentration, orthe like willreflect the dependent variables on which the independent variables of work and heat are controlling.

In certain embodiments of an apparatus in accordance with the invention, and a method 202 in accordance therewith, intervening 206 at levelthree controlmay involve numericalmethods implemented to predict and stab ley step the commands 216 to set points that are expected, projected, predicted, or otherwise calculated to secure propervalues of the dependent variables ofpres sure, temperature, mas s flow rates, and concentration at any particular point within the system 10.

Upon the intervening 206, the system, and particularly the controller 84, may render a decision 207 on whether or not the system 10 is stable. If the system 10 is stable, continuing intervention 206 of the level three control may involve trimming in a predictive fashion any independent variable neces sary. However, if the decision 207 is that the system 10 does not appear to be entirely stable, the pro cess 202 may advance to detecting 209 an event 208 responsible.

For example, if the system 10 does not appear stable, certain events 208 are occurring that may be the effect of atmospheric pres sure, change in concentration in the input flows 35a, orthe like. Any drifting of the system 10 away from the predictive trim control of the intervention 206 will typically be a result of an event 208 altering the condition of the system 10.

Accordingly, detecting 209 the consequences will typically involve feedback 217 from sensors of pres sure, temperature, mass flow rate, concentration, a combination thereof, or the relationships therebetween. For example, the system may encounter a decrease in atmospheric pressure. Likewise, the system 10 may detect an increase in mass flow rate of distillate. In this example, these changes will be detected by sensor 72 and reported back to the controller 84 as data inputs , reflecting consequences of the event 208.

Following detecting 209 these consequences , activation 210 of the control through the co ntro ller 84 is appropriate. The level zero control 203 is left out of the controlloop ofthe proces s 202 for purposes of illustration. The level zero control involves control of parameters that can easily b e observed, controlled, and immediately affected. Adding liquid through adjustment of the rate of flow through a pump 76b will result in increased flow 35a into the tank 30. Likewise, an increase in the speed o f a p ump 76d may o ccur by slaving the control for the pump 76d to the volumetric flow rate, speed, current, or other control mechanism of the pump 76b .

In contrast, determining exactly how much heat should be added to the auxiliary heater 70 is not neces sarily an intuitive proces s and is certainly not directly observable nor controllable manually or by a simple feedback sensor. The time of response for the temperature in the core 20 or the tank 30 is comparatively long (e.g., 4.6 hours), and the responsivenes s of the compres sor 50 is so fast (seconds), that a measurement on a sensor 72d in the plenum 58 does not neces sarily provide an obvious direction for intervention 204, 205 at levels one or two, respectively.

The level zero control may also be trimmed by the intervention 206 at level three. Slight adjustments may be made for los ses, miscalculations , calibrations, and the like. However, as a practical matter, the level zero control need not be included in the control loop of the proces s 202.

One way to considerthe intervention 206 at level three controlis in terms ofpredicting what controlparameters should be adjusted, and in which direction they should be adjusted, based on an algorithmic prediction of where the system needs to move, so to speak. Thus, rather than simply tracking a dependent variable and adjusting a single independent variable, the predictive trim control intervention 206 is very much a sophisticated function reflecting the sophisticated interrelationships between heat and mas s transport within the system 10 and among its many components .

Another way to think of the control proces s 202 is with zero level of control 203 maintaining a mas s balance according to the first law of thermodynamics . The mas s within a system must be the mas s flowing in les s the mas s flowing out. Likewise, the intervention 204 at the levelone controlrepresents an energy balance. That is , modifying 211 the work being input as the primary energy source in the operation of the system 10. That is , heat from the auxiliary heater 70 does not operate the system 10. Powerorworkby the compressor50 operates the system 10 and makes up the energy los ses required by the thermodynamic cycle thereof.

Intervention 205 at the leveltwo control actually is not a principle control of the system 10. Rather, intervening 205 by adjusting 213 auxiliary heat is a mechanism for adjusting the operational parameters of the system 10 in accordance with outside effects .

For example, if a storm front rolls in, then atmospheric pres sure will decrease. Since the tank 30 is not sealed as a pressure ves sel, the pres sure in the plenum 58 may track ambient or atmospheric pres sure. A pres sure drop in the plenum 58 may easily be larger than the temperature differential being controlled above atmospheric in the plenum 58.

Thus , intervening 205 at leveltwo involves adjusting the temperature ofthe tankbrine 23 in order to adjust the overall operation of the system 10 to changing outside conditions . Intervening205 at level two may be resetting the system to adjust to a new steady state of operation within its environment. Environment cannot be controlled by the system. Rather, the system 10 must adjust to its environment and does so by the intervention 205. Therefore, intervention 205 is prospective to the extent that it is initiated as a result of intervening 204 at level one. However, it may still be directed to readjusting the parameters of the system 10, so the system 10 may arrive timely at a new and future equilibrium and steady state condition.

Thus , intervening 206 at level three of control is almost entirely predictive. All the lower levels of control are considered and the operational characteristics are modeled in order to determine the nonobvious set points to which independent variables must be set. Dependent variables thereby arrive at their steady state and proper conditions, in the most effective and timely manner.

Thus , each ofthe levels including intervention 203 at levelzero, intervention 204 at level one, intervention 205 at level two, and intervention 206 at level three abstract the control by the controller 84. Control moves further from direct values of measurable parameters , and away from direct response to current conditions .

Another way to think of this control proces s 202 is with levelzero effectively independent closed loop control of a mass balance directly, direct co ntrol of the value. Intervening 204 at level one is as serting control over an independent variable directly, and the dependent variable indirectly, by a change in work.

Meanwhile, intervening 205 at leveltwo involves addressing the parameters that affect the rate of change and the direction of change, ratherthan affecting the observed variable itself. Finally, intervening 206 at level three involves predicting rates of change of rates of change of parameters to be controlled.

The present invention may be embodied in other specific forms without departing from its fundamental functions oressentialcharacteristics . The described embodiments are to be considered in allrespects only as illustrative, and not restrictive. Allchanges which come within the meaning and range of equivalency ofthe illustrative embodiments are to be embraced within their scope.

Wherefore, we claim: