Fouling of Heat Transfer Surfaces

Transcription

1 20 Fouling o Heat Transer Suraces Mostaa M. Awad Mansoura University, Faculty o Engineering, Mech. Power Eng. Dept., Egypt 1. Introduction Fouling is generally deined as the accumulation and ormation o unwanted materials on the suraces o processing equipment, which can seriously deteriorate the capacity o the surace to transer heat under the temperature dierence conditions or which it was designed. Fouling o heat transer suraces is one o the most important problems in heat transer equipment. Fouling is an extremely complex phenomenon. Fundamentally, ouling may be characterized as a combined, unsteady state, momentum, mass and heat transer problem with chemical, solubility, corrosion and biological processes may also taking place. It has been described as the major unresolved problem in heat transer 1. According to many [1-3], ouling can occur on any luid-solid surace and have other adverse eects besides reduction o heat transer. It has been recognized as a nearly universal problem in design and operation, and it aects the operation o equipment in two ways: Firstly, the ouling layer has a low thermal conductivity. This increases the resistance to heat transer and reduces the eectiveness o heat exchangers. Secondly, as deposition occurs, the cross sectional area is reduced, which causes an increase in pressure drop across the apparatus. In industry, ouling o heat transer suraces has always been a recognized phenomenon, although poorly understood. Fouling o heat transer suraces occurs in most chemical and process industries, including oil reineries, pulp and paper manuacturing, polymer and iber production, desalination, ood processing, dairy industries, power generation and energy recovery. By many, ouling is considered the single most unknown actor in the design o heat exchangers. This situation exists despite the wealth o operating experience accumulated over the years and accumulation o the ouling literature. This lake o understanding almost relects the complex nature o the phenomena by which ouling occurs in industrial equipment. The wide range o the process streams and operating conditions present in industry tends to make most ouling situations unique, thus rendering a general analysis o the problem diicult. In general, the ability to transer heat eiciently remains a central eature o many industrial processes. As a consequence much attention has been paid to improving the understanding o heat transer mechanisms and the development o suitable correlations and techniques that may be applied to the design o heat exchangers. On the other hand relatively little consideration has been given to the problem o surace ouling in heat exchangers. The 1 The unresolved problems in heat transer are: low induced tube vibration, ouling, mixture boiling, low distribution in two-phase low and detailed turbulence low modelling.

2 508 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 506 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems principal purpose o this chapter is to provide some insight into the problem o ouling rom a scientiic and technological standpoint. A better understanding o the problem and o the mechanisms that lead to the accumulation o deposits on suraces will provide opportunities to reduce or even eliminate the problem in certain situations. Fouling can occur as a result o the luids being handled and their constituents in combination with the operating conditions such as temperature and velocity. Almost any solid or semi solid material can become a heat exchanger oulant, but some materials that are commonly encountered in industrial operations as oulants include: Inorganic materials Airborne dusts and grit Waterborne mud and silts Calcium and magnesium salts Iron oxide Organic materials Biological substances, e.g. bacteria, ungi and algae Oils, waxes and greases Heavy organic deposits, e.g. polymers, tars Carbon Energy conservation is oten a actor in the economics o a particular process. At the same time in relation to the remainder o the process equipment, the proportion o capital that is required to install the exchangers is relatively low. It is probably or this reason that heat exchanger ouling has been neglected as most ouling problems are unique to a particular process and heat exchanger design. The problem o heat exchanger ouling thereore represents a challenge to designers, technologists and scientists, not only in terms o heat transer technology but also in the wider aspects o economics and environmental acceptability and the human dimension. 2. Types o ouling Many types o ouling can occur on the heat transer suraces. The generally avored scheme or the classiication o the heat transer ouling is based on the dierent physical and chemical processes involved. Nevertheless, it is convenient to classiy the ouling main types as: 1. Particulate ouling: It is the deposition o suspended particles in the process streams onto the heat transer suraces. I the settling occurs due to gravity as well as other deposition mechanisms, the resulting particulate ouling is called "sedimentation" ouling. Hence, particulate ouling may be deined as the accumulation o particles rom heat exchanger working luids (liquids and/or gaseous suspensions) on the heat transer surace. Most oten, this type o ouling involves deposition o corrosion products dispersed in luids, clay and mineral particles in river water, suspended solids in cooling water, soot particles o incomplete combustion, magnetic particles in economizers, deposition o salts in desalination systems, deposition o dust particles in air coolers, particulates partially present in ire-side (gas-side) ouling o boilers, and so on. The particulate ouling is inluenced by the ollowing actors: concentration o suspended particles, luid low velocity, temperature conditions on the ouled surace (heated or nonheated), and heat lux at the heat transer surace. 2. Crystallization or precipitation ouling: It is the crystallization o dissolved salts rom saturated solutions, due to solubility changes with temperature, and subsequent precipitation onto the heat transer surace. It generally occurs with aqueous solutions and

3 Fouling o Heat Transer Suraces 509 Fouling o Heat Transer Suraces 507 other liquids o soluble salts which are either being heated or cooled. The deposition o inverse solubility salts on heated suraces, usually called "scaling" and its deposited layer is hard and tenacious. The deposition o normal solubility salts on cooled suraces, usually has porous and mushy deposited layers and it is called "sludge", "sotscale", or "powdery deposit". Precipitation/crystallization ouling is common when untreated water, seawater, geothermal water, brine, aqueous solutions o caustic soda, and other salts are used in heat exchangers. The most important phenomena involved with this type o ouling include crystal growth during precipitation require ormation o a primary nucleus. The mechanism controlling that process is nucleation, as a rule heterogeneous in the presence o impurities and on the heat transer surace. 3. Chemical reaction ouling: The deposition in this case is the result o one or more chemical reactions between reactants contained in the lowing luid in which the surace material itsel is not a reactant or participant. However, the heat transer surace may act as a catalyst as in cracking, coking, polymerization, and autoxidation. Thermal instabilities o chemical species, such as asphaltenes and proteins, can also induce ouling precursors. This ouling occurs over a wide temperature range rom ambient to over 1000 o C but is more pronounced at higher temperatures. The mechanism o this type o ouling is a consequence o an unwanted chemical reaction that takes place during the heat transer process. Chemical reaction ouling is ound in many applications o process industry, such as petrochemical industries, oil reining, vapor-phase pyrolysis, cooling o gas and oils, polymerization o process monomers, and so on. Furthermore, ouling o heat transer surace by biological luids may involve complex heterogeneous chemical reactions and physicochemical processes. The deposits rom chemical reaction ouling may promote corrosion at the surace i the ormation o the protective oxide layer is inhibited. 4. Corrosion ouling: It involves a chemical or electrochemical reaction between the heat transer surace itsel and the luid stream to produce corrosion products which, in turn, change the surace thermal characteristics and oul it. Corrosion may cause ouling in two ways. First, corrosion products can accumulate and adhere to the surace providing resistance to heat transer. Second, corrosion products may be transported as particulate materials rom the corrosion site and be deposited as particulate ouling on the heat transer surace in another site o the system. For example, ouling on the water side o boilers may be caused by corrosion products that originate in the condenser or eedtrain. Corrosion ouling is prevalent in many applications where chemical reaction ouling takes place and the protective oxide layer is not ormed on the surace. It is o signiicant importance in the design o the boiler and condenser o a ossil uel ired power plant. 5. Biological ouling: It is the attachment and growth o macroorganisms and /or microorganisms and their products on the heat transer surace. It is usually called "Bioouling", and it is generally a problem in water streams. In general, biological ouling can be divided into two main subtypes o ouling: microbial and macrobial. Microbial ouling is accumulation o microorganisms such as algae, ungi, yeasts, bacteria, and molds, and macrobial ouling represents accumulation o macroorganisms such as clams, barnacles, mussels, and vegetation as ound in seawater or estuarine cooling water. Microbial ouling precedes macrobial deposition as a rule and may be considered o primary interest. Biological ouling is generally in the orm o a bioilm or a slime layer on the surace that is uneven, ilamentous, and deormable but diicult to remove. Although biological ouling

4 510 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 508 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems could occur in suitable liquid streams, it is generally associated with open recirculation or once-through systems with cooling water. Biological ouling may promote corrosion ouling under the slime layer. 6. Solidiication or reezing ouling: It is the reezing o a pure liquid or a higher melting point components o a multicomponent solution onto a subcooled suraces. Separation o waxes rom hot streams when it come to contact with cooled suraces, ormation o ice on a heat transer surace during chilled water production or cooling o moist air, deposits ormed in phenol coolers, and deposits ormed during cooling o mixtures o substances such as parain are some examples o solidiication ouling. This ouling mechanism occurs at low temperatures, usually ambient and below depending on local pressure conditions. The main actors aecting solidiication ouling are mass low rate o the working luid, temperature and crystallization conditions, surace conditions, and concentration o the solid precursor in the luid. It should be noted that, in many applications, where more than one ouling mechanism is present, the ouling problem becomes very complex with their synergistic eects. It is obvious that one cannot talk about a single, uniied theory to model the ouling process wherein not only the oregoing six types o ouling mechanisms are identiied, but in many processes more than one ouling mechanism exists with synergistic eects. 3. Fouling processes The overall ouling process is usually considered to be the net result o two simultaneous sub-processes; a deposition process and a removal (reentrainment) process. A schematic representation o ouling process is given in Fig. (1). All sub-processes can be summarzed as: - Formation o oulant materials in the bulk o the luid. - Transport o oulant materials to the deposit-luid interace. - Attachment/ ormation reaction at the deposit-luid interace. - Removal o the ouling deposit ( spalling or sloughing o the deposit layer). - Transport rom the deposit-luid interace to the bulk o the luid. A schematic diagram or the ouling processes is shown in Fig. (2). It must be noted that, some o these sub-processes may not be applicable in certain ouling situations such as corrosion ouling. Deposition rate ( d ) Flow Stream Removal rate ( r ) Fouling layer Heat Transer Surace Fig. 1. Fouling processes

5 Fouling o Heat Transer Suraces 511 Fouling o Heat Transer Suraces 509 Fouling Deposition Process Formation in the bulk o the luid Transport to the deposit-luid interace Attachment/ ormation reaction at the deposit-luid nterace Removal Process Removal o the ouling deposit Transport rom the deposit-luid interace Fig. 2. Schematic diagram or the ouling processes In another way, three basic stages may be visualized in relation to deposition on suraces rom a moving luid. They are: 1. The diusional transport o the oulant or its precursors across the boundary layers adjacent to the solid surace within the lowing luid. 2. The adhesion o the deposit to the surace and to itsel. 3. The transport o material away rom the surace. The sum o these basic components represents the growth o the deposit on the surace. In mathematical terms the rate o' deposit growth (ouling resistance or ouling actor, R ) may be regarded as the dierence between the deposition and removal rates as: R d r (1) where d and r are the rates o deposition and removal respectively. The ouling actor, R, as well as the deposition rate, d, and the removal rate, r, can be expressed in the units o thermal resistance as m 2 K/W or in the units o the rate o thickness change as m/s or units o mass change as kg/ m 2 s. 4. Deposition and removal mechanisms From the empirical evidence involving various ouling mechanisms discussed in Section 2, it is clear that virtually all these mechanisms are characterized by a similar sequence o events. The successive events occurring in most cases are illustrated in Fig. (2). These events govern the overall ouling process and determine its ultimate impact on heat exchanger perormance. In some cases, certain events dominate the ouling process, and they have a direct eect on the type o ouling to be sustained. The main ive events can be summarized briely as ollowing:

6 512 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 510 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 1-Formation o oulant materials in the bulk o the luid or initiation o the ouling, the irst event in the ouling process, is preceded by a delay period or induction period, t d as shown in Fig. (3), the basic mechanism involved during this period is heterogeneous nucleation, and t d is shorter with a higher nucleation rate. The actors aecting t d are temperature, luid velocity, composition o the ouling stream, and nature and condition o the heat exchanger surace. Low-energy suraces (unwettable) exhibit longer induction periods than those o high-energy suraces (wettable). In crystallization ouling, t d tends to decrease with increasing degree o supersaturation. In chemical reaction ouling, t d appears to decrease with increasing surace temperature. In all ouling mechanisms, t d decreases as the surace roughness increases due to available suitable sites or nucleation, adsorption, and adhesion. 2-Transport o species means transer o the ouling species itsel rom the bulk o the luid to the heat transer surace. Transport o species is the best understood o all sequential events. Transport o species takes place through the action o one or more o the ollowing mechanisms: Diusion: involves mass transer o the ouling constituents rom the lowing luid toward the heat transer surace due to the concentration dierence between the bulk o the luid and the luid adjacent to the surace. Electrophoresis: under the action o electric orces, ouling particles carrying an electric charge may move toward or away rom a charged surace depending on the polarity o the surace and the particles. Deposition due to electrophoresis increases with decreasing electrical conductivity o the luid, increasing luid temperature, and increasing luid velocity. It also depends on the ph o the solution. Surace orces such as London van der Waals and electric double layer interaction orces are usually responsible or electrophoretic eects. Thermophoresis: a phenomenon whereby a "thermal orce" moves ine particles in the direction o negative temperature gradient, rom a hot zone to a cold zone. Thus, a high-temperature gradient near a hot wall will prevent particles rom depositing, but the same absolute value o the gradient near a cold wall will promote particle deposition. The thermophoretic eect is larger or gases than or liquids. Diusiophoresis: involves condensation o gaseous streams onto a surace. Sedimentation: involves the deposition o particulate matters such as rust particles, clay, and dust on the surace due to the action o gravity. For sedimentation to occur, the downward gravitational orce must be greater than the upward drag orce. Sedimentation is important or large particles and low luid velocities. It is requently observed in cooling tower waters and other industrial processes where rust and dust particles may act as catalysts and/or enter complex reactions. Inertial impaction: a phenomenon whereby large particles can have suicient inertia that they are unable to ollow luid streamlines and as a result, deposit on the surace. Turbulent downsweeps: since the viscous sublayer in a turbulent boundary layer is not truly steady, the luid is being transported toward the surace by turbulent downsweeps. These may be thought o as suction areas o measurable strength distributed randomly all over the surace. 3-Attachment o the ouling species to the surace involves both physical and chemical processes, and it is not well understood. Three interrelated actors play a crucial role in the attachment process: surace conditions, surace orces, and sticking probability. It is the combined and simultaneous action o these actors that largely accounts or the event o attachment.

7 Fouling o Heat Transer Suraces 513 Fouling o Heat Transer Suraces 511 Surace properties: The properties o surace conditions important or attachment are the surace ree energy, wettability (contact angle, spreadability), and heat o immersion. Wettability and heat o immersion increase as the dierence between the surace ree energy o the wall and the adjacent luid layer increases. Unwettable or low-energy suraces have longer induction periods than wettable or high-energy suraces, and suer less rom deposition (such as polymer and ceramic coatings). Surace roughness increases the eective contact area o a surace and provides suitable sites or nucleation and promotes initiation o ouling. Hence, roughness increases the wettability o wettable suraces and decreases the unwettability o the unwettable ones. Surace orces: The most important one is the London van der Waals orce, which describes the intermolecular attraction between nonpolar molecules and is always attractive. The electric double layer interaction orce can be attractive or repulsive. Viscous hydrodynamic orce inluences the attachment o a particle moving to the wall, which increases as it moves normal to the plain surace. Sticking probability: represents the raction o particles that reach the wall and stay there beore any reentrainment occurs. It is a useul statistical concept devised to analyze and explain the complicated event o attachment. 4-Removal o the ouling deposits rom the surace may or may not occur simultaneously with deposition. Removal occurs due to the single or simultaneous action o the ollowing mechanisms; shear orces, turbulent bursts, re-solution, and erosion. Shear orces result rom the action o the shear stress exerted by the lowing luid on the depositing layer. As the ouling deposit builds up, the cross-sectional area or low decreases, thus causing an increase in the average velocity o the luid or a constant mass low rate and increasing the shear stress. Fresh deposits will orm only i the deposit bond resistance is greater than the prevailing shear orces at the solid luid interace. Randomly distributed (about less than 0.5% at any instant o time) periodic turbulent bursts act as miniature tornadoes liting deposited material rom the surace. By continuity, these luid bursts are compensated or by gentler luid back sweeps, which promote deposition. Re-solution: The removal o the deposits by re-solution is related directly to the solubility o the material deposited. Since the ouling deposit is presumably insoluble at the time o its ormation, dissolution will occur only i there is a change in the properties o the deposit, or in the lowing luid, or in both, due to local changes in temperature, velocity, alkalinity, and other operational variables. For example, suiciently high or low temperatures could kill a biological deposit, thus weakening its attachment to a surace and causing sloughing or re-solution. The removal o corrosion deposits in power-generating systems is done by re-solution at low alkalinity. Resolution is associated with the removal o material in ionic or molecular orm. Erosion is closely identiied with the overall removal process. It is highly dependent on the shear strength o the oulant and on the steepness and length o the sloping heat exchanger suraces, i any. Erosion is associated with the removal o material in particulate orm. The removal mechanism becomes largely ineective i the ouling layer is composed o well-crystallized pure material (strong ormations); but it is very eective i it is composed o a large variety o salts each having dierent crystal properties.

8 514 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 512 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 5- Transport rom the deposit-luid interace to the bulk o the luid, once the deposits are sloughed, it may/may not transported rom the deposit-luid interace to the bulk o the luid. This depend on the mass and volume o the sloughed piece and on the hydrodynamic orces o the lowing luid. I the sloughed piece is larg enough, it may moved on the surace and depoited on another site on the system such as some corrosion products. All deposits which removed due to erosion eect will be transported to the bulk o the luid. The removal process in not complete without this action. The important parameter aecting the deposit sloughing is the aging o deposits in which it may strengthen or weaken the ouling deposits. 5. Fouling curves The overall process o ouling is indicated by the ouling actor, R (ouling resistance) which is measured either by a test section or evaluated rom the decreased capacity o an operating heat exchanger. The representation o various modes o ouling with reerence to time is known as a ouling curve (ouling actor-time curve). Typical ouling curves are shown in Fig. (3). R Linear Falling R * Asymptotic d = r t d t * t c Time, t Fig. 3. Fouling Curves The delay time, t d indicates that an initial period o time can elapse where no ouling occurs. The value o t d is not predictable, but or a given surace and system, it appears to be somewhat random in nature or having a normal distribution about some mean value or at least dependent upon some requency actors. Ater clean the ouled suraces and reused them, the delay time, t d is usually shorter than that o the new suraces when are used or the irst time. It must be noted that, the nature o ouling actor-time curve is not a unction o t d. The most important ouling curves are: - Linear ouling curve is indicative o either a constant deposition rate, d with removal rate, r being negligible (i.e. d = constant, r 0) or the dierence between d and r

9 Fouling o Heat Transer Suraces 515 Fouling o Heat Transer Suraces 513 is constant (i.e. d r = constant). In this mode, the mass o deposits increases gradually with time and it has a straight line relationship o the orm (R = at) where a is the slope o the line. - Falling rate ouling curve results rom either decreasing deposition rate, d with removal rate, r being constant or decreasing deposition rate, d and increasing removal rate, r. In this mode, the mass o deposit increases with time but not linearly and does not reach the steady state o asymptotic value. - Asymptotic ouling curve is indicative o a constant deposition rate, d and the removal rate, r being directly proportional to the deposit thickness until d = r at the asymptote. In this mode, the rate o ouling gradually alls with time, so that eventually a steady state is reached when there is no net increase o deposition on the surace and there is a possibility o continued operation o the equipments without additional ouling. In practical industrial situations, the asymptote may be reached and the asymptotic ouling actor, R * is obtained in a matter o minutes or it may take weeks or months to occur depending on the operating conditions. The general equation describing this behavior is given in equation (4). This mode is the most important one in which it is widely existed in the industrial applications. The pure particulate ouling is one o this type. For all ouling modes, the amount o material deposited per unit area, m is related to the ouling resistance (R ), the density o the oulant ( ), the thermal conductivity ( ) and the thickness o the deposit (x ) by the ollowing equation: where m x R (2) R x (3) (values o thermal conductivities or some oulants are given in table 1). Foulant Alumina Bioilm (eectively water) Carbon Calcium sulphate Calcium carbonate Magnesium carbonate Titanium oxide Wax Table 1. Thermal conductivities o some oulants [2] Thermal conductivity (W/mK) It should be noted that, the curves represented in Fig. (3) are ideal ones while in the industrial situations, ideality may not be achieved. A closer representation o asymptotic ouling practical curve might be as shown in Fig. (4). The saw tooth eect is the result o partial removal o some deposit due to spalling or sloughing to be ollowed or a short

10 516 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 514 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems time by a rapid build up o deposit. The average curve (represented by the dashed line) can be seen to represent the ideal asymptotic curve on Fig. (3). Similar eects o partial removal and deposition may be experienced with the other types o oulin curves. Fouling resistance, ( R) t d Time, (t) Fig. 4. Practical ouling curve 6. Cost o ouling Fouling aects both capital and operating costs o heat exchangers. The extra surace area required due to ouling in the design o heat exchangers, can be quite substantial. Attempts have been made to make estimates o the overall costs o ouling in terms o particular processes or in particular countries. Reliable knowledge o ouling economics is important when evaluating the cost eiciency o various mitigation strategies. The total ouling-related costs can be broken down into our main areas: 4. Higher capital expenditures or oversized plants which includes excess surace area (10-50%), costs or extra space, increased transport and installation costs. 5. Energy losses due to the decrease in thermal eiciency and increase in the pressure drop. 6. Production losses during planned and unplanned plant shutdowns or ouling cleaning. 7. Maintenance including cleaning o heat transer equipment and use o antioulants. The loss o heat transer eiciency usually means that somewhere else in the system, additional energy is required to make up or the short all. The increased pressure drop through a heat exchanger represents an increase in the pumping energy required to maintain the same low rate. The ouling resistance used in any design brings about 50% increase in the surace area over that required i there is no ouling. The need or additional maintenance as a result o ouling may be maniested in dierent ways. In general, any extensive ouling means that the heat exchanger will have to be cleaned on a regular basis to restore the loss o its heat transer capacity. According to Pritchard [4], the total heat exchanger ouling costs or highly industrialized countries are about 0.25% o the countries' Gross National Product (GNP). Table (2) shows the annual costs o ouling in some dierent countries based on 1992 estimation.

11 Fouling o Heat Transer Suraces 517 Fouling o Heat Transer Suraces 515 Country Fouling Costs (million $) Fouling Cost /GNP % US UK Germany France Japan Australia New Zealand Table 2. Annual costs o ouling in some countries (1992 estimation) [5]. From this table, it is clear that ouling costs are substantial and any reduction in these costs would be a welcome contribution to proitability and competitiveness. The requency o cleaning will o course depend upon the severity o the ouling problem and may range between one weak and one year or longer. Frequent cleaning involving repeated dismantling and reassembly will inevitably result in damage to the heat exchanger at a lesser or greater degree, which could shorten the useul lie o the equipment. Fouling can be very costly in reinery and petrochemical plants since it increases uel usage, results in interrupted operation and production losses, and increases maintenance costs. Increased Capital Investment In order to make allowance or potential ouling the area or a given heat transer surace is larger than or clean conditions. To accommodate the ouling-related drop in heat transer capacity, the tubular exchangers are generally designed with 20-50% excess surace, where the compact heat exchangers are designed with 15-25% excess surace. In addition to the actual size o the heat exchanger other increased capital costs are likely. For instance where it is anticipated that a particular heat exchanger is likely to suer severe or diicult ouling, provision or o-line cleaning will be required. The location o the heat exchanger or easy access or cleaning may require additional pipe work and larger pumps compared with a similar heat exchanger operating with little or no ouling placed at a more convenient location. Furthermore i the problem o ouling is thought to be excessive it might be necessary to install a standby exchanger, with all the associated pipe work oundations and supports, so that one heat exchanger can be operated while the other is being cleaned and serviced. Under these circumstances the additional capital cost is likely to more than double and with allowances or heavy deposits the inal cost could be 4-8 times the cost o the corresponding exchanger running in a clean condition. Additional capital costs may be considered or on-line cleaning such as the Taprogge system (see sec. 12) or other systems. It has to be said however, that on-line cleaning can be very eective and that the additional capital cost can oten be justiied in terms o reduced operating costs. Furthermore the way in which the additional area is accommodated, can aect the rate o ouling. For instance i the additional area results say, in reduced velocities, the ouling rate may be higher than anticipated and the value o the additional area may be largely oset by the eects o heavy deposits. The indiscriminate use o excess surace area or instance, can lead to high capital costs, especially where exotic and expensive materials o construction are required.

12 518 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 516 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems Additional Operating Costs The presence o ouling on the surace o heat exchangers decreases the ability o the unit to transer heat. Due to this decrement in the exchanger thermal capacity, neither the hot stream nor the cold stream will approach its target temperature. To compensate this shortage in the heat low, either additional cooling utility or additional heating utility is required. On the other hand, the presence o deposits on the surace o heat exchangers increases the pressure drop and to recover this increment, an additional pumping work is required and hence a greater pumping cost. Also the ouling may be the cause o additional maintenance costs. The more obvious result o course, is the need to clean the heat exchanger to return it to eicient operation. Not only will this involve labour costs but it may require large quantities o cleaning chemicals and there may be eluent problems to be overcome that add to the cost. I the cleaning agents are hazardous or toxic, elaborate saety precautions with attendant costs, may be required. The requent need to dismantle and clean a heat exchanger can aect the continued integrity o the equipment, i.e. components in shell and tube exchangers such as bales and tubes may be damaged or the gaskets and plates in plate heat exchangers may become aulty. The damage may also aggravate the ouling problem by causing restrictions to low and upsetting the required temperature distribution. Loss o Production The need to restore low and heat exchanger eiciency will necessitate cleaning. On a planned basis the interruptions to production may be minimized but even so i the remainder o the plant is operating correctly then this will constitute a loss o output that, i the remainder o the equipment is running to capacity still represents a loss o proit and a reduced contribution to the overall costs o the particular site. The consequences o enorced shutdown due to the eects o ouling are o course much more expensive in terms o output. Much depends on recognition o the potential ouling at the design stage so that a proper allowance is made to accommodate a satisactory cleaning cycle. When the seriousness o a ouling problem goes unrecognized during design then unscheduled or even emergency shutdown, may be necessary. Production time lost through the need to clean a heat exchanger can never be recovered and it could in certain situations, mean the dierence between proit and loss. The Cost o Remedial Action I the ouling problem cannot be relieved by the use o additives it may be necessary to make modiications to the plant. Modiication to allow on-line cleaning o a heat exchanger can represent a considerable capital investment. Beore capital can be committed in this way, some assessment o the eectiveness o the modiication must be made. In some examples o severe ouling problems the decision is straightorward, and a pay back time o less than a year could be anticipated. In other examples the decision is more complex and the inancial risks involved in making the modiication will have to be addressed. A number o contributions to the cost o ouling have been identiied, however some o the costs will remain hidden. Although the cost o cleaning and loss o production may be recognized and properly assessed, some o the associated costs may not be attributed directly to the ouling problem. For instance the cost o additional maintenance o ancillary equipment such as pumps and pipework, will usually be lost in the overall maintenance charges.

13 Fouling o Heat Transer Suraces 519 Fouling o Heat Transer Suraces Parameters aecting ouling The ouling process is a dynamic and unsteady one in which many operational and design variables have been identiied as having most pronounced and well deined eects on ouling. These variables are reviewed in principle to clariy the ouling problems and because the designer has an inluence on their modiication. Those parameters include the luid low velocity, the luid properties, the surace temperature, the surace geometry, the surace material, the surace roughness, the suspended particles concentration and properties,.etc. According to many investigators, the most important parameters are: 1. Fluid low velocity The low velocity has a strong eect on the ouling rate where it has direct eects on both o the deposition and removal rates through the hydrodynamic eects such as the eddies and shear stress at the surace. On the other hand, the low velocity has indirect eects on deposit strength ( ), the mass transer coeicient (k m ), and the stickability (P). It is well established that, increasing the low velocity tends to increase the thermal perormance o the exchanger and decrease the ouling rate. Uniorm and constant low o process luids past the heat transer surace avors less ouling. Foulants suspended in the process luids will deposit in low-velocity regions, particularly where the velocity changes quickly, as in heat exchanger water boxes and on the shell side. Higher shear stress promotes dislodging o deposits rom suraces. Maintain relatively uniorm velocities across the heat exchanger to reduce the incidence o sedimentation and accumulation o deposits. 2. Surace temperature The eect o surace temperature on the ouling rate has been mentioned in several studies. These studies indicated that the role o surace temperature is not well deined. The literatures show that, "increase surace temperature may increase, decrease, or has no eect on the ouling rates". This variation in behavior does indicate the importance to improve our understanding about the eect o surace temperature on the ouling process, A good practical rule to ollow is to expect more ouling as the temperature rises. This is due to a baking on eect, scaling tendencies, increased corrosion rate, aster reactions, crystal ormation and polymerization, and loss in activity by some antioulants [6]. Lower temperatures produce slower ouling buildup, and usually deposits that are easily removable [7]. However, or some process luids, low surace temperature promotes crystallization and solidiication ouling. To overcome these problems, there is an optimum surace temperature which better to use or each situation. For cooling water with a potential to scaling, the desired maximum surace temperature is about 60 C. Biological ouling is a strong unction o temperature. At higher temperatures, chemical and enzyme reactions proceed at a higher rate with a consequent increase in cell growth rate [8]. According to Mukherjee [8], or any biological organism, there is a temperature below which reproduction and growth rate are arrested and a temperature above which the organism becomes damaged or killed. I, however, the temperature rises to an even higher level, some heat sensitive cells may die. 3. Surace material The selection o surace material is signiicant to deal with corrosion ouling. Carbon steel is corrosive but least expensive. Copper exhibits biocidal eects in water. However, its use is limited in certain applications: (1) Copper is attacked by biological organisms including

14 520 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 518 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems sulate-reducing bacteria; this increases ouling. (2) Copper alloys are prohibited in highpressure steam power plant heat exchangers, since the corrosion deposits o copper alloys are transported and deposited in high-pressure steam generators and subsequently block the turbine blades. (3) Environmental protection limits the use o copper in river, lake, and ocean waters, since copper is poisonous to aquatic lie. Noncorrosive materials such as titanium and nickel will prevent corrosion, but they are expensive and have no biocidal eects. Glass, graphite, and telon tubes oten resist ouling and/or improve cleaning but they have low thermal conductivity. Although the construction material is more important to resist ouling, surace treatment by plastics, vitreous enamel, glass, and some polymers will minimize the accumulation o deposits. 4. Surace Roughness The surace roughness is supposed to have the ollowing eects: (1) The provision o nucleation sites that encourage the laying down o the initial deposits. (2) The creation o turbulence eects within the lowing luid and, probably, instabilities in the viscous sublayer. Better surace inish has been shown to inluence the delay o ouling and ease cleaning. Similarly, non-wetting suraces delay ouling. Rough suraces encourage particulate deposition and provide a good chance or deposit sticking. Ater the initiation o ouling, the persistence o the roughness eects will be more a unction o the deposit itsel. Even smooth suraces may become rough in due course due to scale ormation, ormation o corrosion products, or erosion. 5. Fluid Properties The luid propensity or ouling is depending on its properties such as viscosity and density. The viscosity is playing an important rule or the sublayer thickness where the deposition process is taking place. On the other side the viscosity and density have a strong eect on the sheer stress which is the key element in the removal process. Fig. 5. Eect o the low luid type on the ouling To show the eect o the low luid type on the ouling resistance, Chenoweth [7} collected data rom over 700 shell and-tube heat exchangers. These data o combined shell- and tubeside ouling resistances (by summing each side entry), have been compiled and divided into nine combinations o liquid, two-phase, and gas on each luid side regardless o the applications. The arithmetic average o total R o each two-luid combination value has been taken and analyzed. The results are presented in Fig. (5) with ordinate ranges between 0 and

15 Fouling o Heat Transer Suraces 521 Fouling o Heat Transer Suraces From this igure, it is clear that the maximum value is 1.0, that is due to liquid-liquid heat exchanger, where the minimum value is 0.5 which belong to gas-gas heat exchanger. I liquid is on the shell side and gas on the tube side, the relative ouling resistance is However, i liquid is on the tube side and gas on the shell side, it is Since many process industry applications deal with liquids that are dirtier than gases, the general practice is to speciy larger ouling resistances or liquids compared to those or the gases. Also, i ouling is anticipated on the liquid side o a liquid gas exchanger, it is generally placed in the tubes or cleaning purposes spite a larger ouling resistance is speciied. These trends are clear rom the igure. It should again be emphasized that Fig. (5) indicates the current practice and has no scientiic basis. Speciication o larger ouling resistances or liquids (which have higher heat transer coeicients than those o gases) has even more impact on the surace area requirement or liquid liquid exchangers than or gas gas exchangers. 6. Impurities and Suspended Solids Seldom are luids pure. Intrusion o minute amounts o impurities can initiate or substantially increase ouling. They can either deposit as a ouling layer or acts as catalysts to the ouling processes [6]. For example, chemical reaction ouling or polymerization o reinery hydrocarbon streams is due to oxygen ingress and/or trace elements such as Va and Mo. In crystallization ouling, the presence o small particles o impurities may initiate the deposition process by seeding. The properties o the impurities orm the basis o many antioulant chemicals. Sometimes impurities such as sand or other suspended particles in cooling water may have a scouring action, which will reduce or remove deposits [9]. Suspended solids promote particulate ouling by sedimentation or settling under gravitation onto the heat transer suraces. Since particulate ouling is velocity dependent, prevention is achieved i stagnant areas are avoided. For water, high velocities (above 1 m/s) help prevent particulate ouling. Oten it is economical to install an upstream iltration. 7. Heat Transer Process The ouling resistances or the same luid can be considerably dierent depending upon whether heat is being transerred through sensible heating or cooling, boiling, or condensing. 8. Design Considerations Equipment design can contribute to increase or decrease ouling. Heat exchanger tubes that extend beyond tube sheet, or example, can cause rapid ouling. Some ouling aspects must be considered through out the equipment design such as: 1. Placing the More Fouling Fluid on the Tube Side As a general guideline, the ouling luid is preerably placed on the tube side or ease o cleaning. Also, there is less probability or low-velocity or stagnant regions on the tube side. 2. Shell-Side Flow Velocities Velocities are generally lower on the shell side than on the tube side, less uniorm throughout the bundle, and limited by low-induced vibration. Zero-or low-velocity regions on the shell side serve as ideal locations or the accumulation o oulants. I ouling is expected on the shell side, then attention should be paid to the selection o bale design. Segmental bales have the tendency or poor low distribution i spacing or bale cut ratio is not in correct proportions. Too low or too high a ratio results in an unavorable low regime that avors ouling.

16 522 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 520 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 3. Low-Finned Tube Heat Exchanger There is a general apprehension that low Reynolds number low heat exchangers with lowinned tubes will be more susceptible to ouling than plain tubes. Fouling is o little concern or inned suraces operating with moderately clean gases. Fin type does not aect the ouling rate, but the ouling pattern is aected or waste heat recovery exchangers. Plain and serrated in modules with identical densities and heights have the same ouling thickness increases in the same period o time. 4. Gasketed Plate Heat Exchangers High turbulence, absence o stagnant areas, uniorm luid low, and the smooth plate surace reduce ouling and the need or requent cleaning. Hence the ouling actors required in plate heat exchangers are normally 10-25% o those used in shell and tube heat exchangers. 5. Spiral Plate Exchangers High turbulence and scrubbing action minimize ouling on the spiral plate exchanger. This permits the use o low ouling actors. 6. Seasonal temperature changes When cooling tower water is used as coolant, considerations are to be given or winter conditions where the ambient temperature may be near zero or below zero on the Celsius scale. The increased temperature driving orce during the cold season contributes to more substantial overdesign and hence over perormance problems, unless a control mechanism has been instituted to vary the water/air low rate as per the ambient temperature. Also the bulk temperature o the cooling water that used in power condensers is changed seasonally. This change inluences the ouling rate to some extent. 8. Fouling measurements and monitoring The ouling resistances can be measured either experimentally or analytically. The main measuring methods include; 1-Direct weighing; the simplest method or assessing the extent o deposition on test suraces in the laboratory is by direct weighing. The method requires an accurate balance so that relatively small changes in deposit mass may be detected. It may be necessary to use thin walled tube to reduce the tare mass so as to increase the accuracy o the method. 2-Thickness measurement; In many examples o ouling the thickness o the deposit is relatively small, perhaps less than 50 m, so that direct measurement is not easy to obtain. A relatively simple technique provided there is reasonable access to the deposit, is to measure the thickness. Using a removable coupon or plate the thickness o a hard deposit such as a scale, may be made by the use o a micrometer or travelling microscope. For a deormable deposit containing a large proportion o water, e.g. a bioilm it is possible to use an electrical conductivity technique 3-Heat transer measurements; In this method, the ouling resistance can be determined rom the changes in heat transer during the deposition process. The basis or subsequent operations will be Equation (14). The data may be reported in terms o changes in overall heat transer coeicient. A major assumption in this method is that the presence o the deposit does not aect the hydrodynamics o the lowing luid. However, in the irst stages o deposition, the surace o the deposit is usually rougher than the metal surace so that the turbulence within the luid is greater than when it is lowing over a smooth surace. As a result the ouling resistance calculated rom the data will be lower than i the increased level o turbulence had been taken into account. It is possible that the increased turbulence osets

17 Fouling o Heat Transer Suraces 523 Fouling o Heat Transer Suraces 521 the thermal resistance o the deposit and negative values o thermal resistance will be calculated. 4-Pressure drop; As an alternative to direct heat transer measurements it is possible to use changes in pressure drop brought about by the presence o the deposit. The pressure drop is increased or a given low rate by virtue o the reduced low area in the ouled condition and the rough character o the deposit. The shape o the curve relating pressure drop with time will in general, ollow an asymptotic shape so that the time to reach the asymptotic ouling resistance may be determined. The method is oten combined with the direct measurement o thickness o the deposit layer. Changes in riction actor may also be used as an indication o ouling o a low channel. 5-Other techniques or ouling assessment; In terms o their eect on heat exchanger perormance the measurement o heat transer reduction or increase in pressure drop provide a direct indication. The simple methods o measuring deposit thickness described earlier are useul, but in general they require that the experiment is terminated so as to provide access to the test sections. Ideally non-intrusive techniques would allow deposition to continue while the experimental conditions are maintained without disturbance. Such techniques include the use o radioactive tracers and optical methods. Laser techniques can be used to investigate the accumulation and removal o deposits. Also, inra red systems are used to investigate the development and removal o bioilms rom tubular test sections. Microscopic examination o deposits may provide some urther evidence o the mechanisms o ouling, but this is generally a "back up" system rather than to give quantitative data. Gas-Side Fouling Measuring Devices The gas-side ouling measuring devices can be classiied into ive groups: heat lux meters, mass accumulation probes, optical devices, deposition probes, and acid condensation probes. A heat lux meter uses the local heat transer per unit area to monitor the ouling. The decrease in heat lux as a unction o time is thus a measure o the ouling buildup. A mass accumulation device measures the ouling deposit under controlled conditions. Optical measuring devices use optical method to determine the deposition rate. Deposition probes are used to measure the deposit thickness. Acid condensation probes are used to collect liquid acid that accumulates on a surace that is at a temperature below the acid dew point o the gas stream. Instruments or Monitoring o Fouling Instruments have been developed to monitor conditions on a tube surace to indicate accumulation o ouling deposits and, in some cases, to indicate the eect on heat exchanger perormance. The ollowing is a summary o the dierent ouling monitors [10, 11]: 1. Removable sections o the ouled surace, which may be used or microscopic examination, mass measurements, and chemical and biological analysis o the deposits. 2. Increase in pressure drop across the heat exchanger length. This method provides a measure o luid rictional resistance, which usually increases with buildup o ouling deposits. This device is relatively inexpensive and is easy to operate. 3. Thermal resistance monitors, which are used to determine the eect o the deposit on overall heat transer resistance. The thermal method o monitoring has the advantage over the others o giving directly inormation that is required or predicting or assessing heat transer perormance.

18 524 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 522 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 9. Perormance data analysis As mentioned above, ouling has many eects on the heat exchanger perornance. It decreases the exchanger thermal capacity and increases the pressure drop through the exchanger as shown in Fig. (6). From the igure it is clear that the total thermal resistance to heat transer is decreased during the irst stages o ouling due to the surace roughness resulting rom initial deposition. Ater that and with deposits building up, the thermal resistace returns to increase again. Fig. 6. Fouling eects on exchanger perormance In order to model and predict the industrial processes ouling problems it is irst necessary to understand what is happening and what are the causes and eects o ouling. To achieve this, it is necessary to careully examine and evaluate all the data and operating conditions at various plants in order to understand what the variables which are eective on ouling and what are the mechanisms o such phenomena. The objective o these eorts will be always to minimize the ouling clean-up / remediation shut-down requency o the plants and to reduce the cost by making the minimum modiication in the processes. The possibility o whether the ouling material is a part o the eed to the system or it is a product o reaction / aggregation / locculation in the system must be clariied. The role o various operating conditions in the system on ouling (pressures, temperatures, compositions, low rates, etc. and their variations) must be understood and quantiied. Only with appropriate modeling considering all the possible driving orces and mechanisms o ouling one may be able to predict the nature o ouling in each case and develop mitigation techniques to combat that. The available ouling history data would be useul to test the packages which will be developed. Considering the diversity o the data, care must be taken in their analysis or any universality conclusions. However, in order to make comparisons between ouling data rom various plants and test the accuracy o the developed packages, it will be necessary to acquire the compositions data o the eed in each plant as well as characteristics and conditions o operations o the process system used in those plants. Only then one can test the accuracy o the models developed and understand why in one case there is ouling and no ouling in another case. Empirical data or ouling resistances have been obtained over many decades by industry since its irst compilation by TEMA in 1941 or shell-and-tube heat exchangers. TEMA ouling resistances [12] are supposed to be representative values, asymptotic values, or those maniested just beore cleaning to be perormed.

19 Fouling o Heat Transer Suraces 525 Fouling o Heat Transer Suraces 523 It should be reiterated that the recommended ouling resistances are believed to represent typical ouling resistances or design. Consequently, sound engineering judgment has to be made or each selection o ouling resistances, keeping in mind that actual values o ouling resistances in any application can be either higher or lower than the resistances calculated. Finally, it must be clear that ouling resistances, although recommended ollowing the empirical data and a sound model, are still constant, independent o time, while ouling is a transient phenomenon. Hence, the value o R selected represents a correct value only at one speciic time in the exchanger operation. Thereore, it needs to be emphasized that the tables may not provide the applicable values or a particular design. They are only intended to provide guidance when values rom direct experience are unavailable. With the use o inite ouling resistance, the overall U value is reduced, resulting in a larger surace area requirement, larger low area, and reduced low velocity which inevitably results in increased ouling. Thus, allowing more surace area or ouling in a clean exchanger may accelerate ouling initially. Typical ouling resistances are roughly 10 times lower in plate heat exchangers (PHEs) than in shell-and-tube heat exchangers (TEMA values), (see Table 3). Process Fluid Sot water Cooling tower water Seawater River water Lube oil Organic solvents Steam (oil bearing) R, (m 2 K/kW) PHEs TEMA Table 3. Liquid-Side Fouling Resistances or PHEs vs. TEMA Values (rom Re.13) 10. Fouling models Fouling is usually considered to be the net result o two simultaneous processes: a deposition process and a removal process. A schematic representation o ouling process is given in Fig. (1). Mathematically, the net rate o ouling can be expressed as the dierence between the deposition and removal rates as given in equation (1). Many attempts have been made to model the ouling process. One o the earliest models o ouling was that by Kern and Seaton [14]. In this model, it was assumed that the rate o deposition mass, m d, remained constant with time t but that the rate o removal mass, m r, was proportional to the accumulated mass, m, and thereore increased with time to approach m d asymptotically. Thus Rate o accumulation = Rate o deposition Rate o removal dm /dt = m = m d - m r (4) then integration o Eqn. (4) rom the initial condition m = 0 at t = 0 gives * t m m (1 e ) (5)

20 526 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems 524 Heat Transer - Theoretical Analysis, Experimental Investigations and Industrial Systems where m * is the asymptotic value o m and = 1/t c. The time constant t c represents the average residence time or an element o ouling material at the heat transer surace. Reerring to Eqn. (2), Eqn. (5) can be expressed in terms o ouling resistance R at time t in terms o the asymptotic value R * by * t R R (1 e ) (6) It is obvious that the real solution would be to ind expressions or R * and t c as a unction o variables aecting the ouling process. The purpose o any ouling model is to assist the designer or indeed the operator o heat exchangers, to make an assessment o the impact o ouling on heat exchanger perormance given certain operating conditions. Ideally a mathematical interpretation o Eqn. (6) would provide the basis or such an assessment but the inclusion o an extensive set o conditions into one mathematical model would be at best, diicult and even impossible. Modeling eorts to produce a mathematical model or ouling process have been based on the general material balance given in Eqn. (4) and centered on evaluating the unctions m d and m r or speciic ouling situations, some o these models are: Watkinson Model: Watkinson [15] reported the eect o luid velocity on the asymptotic ouling resistance in three cases as; 1. Calcium carbonate scaling (with constant surace temperature and constant composition) 2. Gas oil ouling (with constant heat lux) 3. Sand deposition rom water (with constant heat lux) R * = 0.101/(v 1.33 D 0.23 ) (7) R * = 0.55/v 2 (8) R * = 0.015/v 1.2 (9) where; R * the asymptotic ouling resistance v the luid velocity D the tube diameter Taborek, et al. Model: Taborek, et al [16] introduced a water characterization actor to the deposition term to account or the eect o water quality. The deposition term, also involves two processes; (1) Diusion o the potential depositing substance to the surace and (2) Bonding at the surace. They expressed the deposition rate in an arrhenius type equation as the ollowing: where n Ea d kp 1 d exp( ) (10) RT g s