COVER STORY. By Michael L. Dillon, seepex, Inc.

Transcription

1 COVER STORY Maximize the Performance of Progreive Cavity Pump If you chooe the right geometry, protect your part, conider the application carefully, control temperature and preure, your PC pump could lat a long, long time. By Michael L. Dillon, eepex, Inc. A poitive diplacement pump, progreive cavity (PC) pump have the ame benefit a other PD pump. They can handle high vicoity fluid; they can produce accurate repeatable flow; the output capacity i relatively independent of head; and they can operate with fairly high efficiency at high head. One of the major progreive cavity pump benefit i that they have no valve. A PC pump work like a piton pump, but with the piton operating in a cylinder of infinite length. Without valve, the pump will not air lock, clog, foul or leak, and flow i predictable and repeatable. The abence of any valve on the uction ide of the PC pump contribute to it low NPSHR. The PC pump i alo excellent for abraive lurrie, ince in it mot common form, it i rubber lined. It can be repaired on-ite with only a few tolerance fit. It can handle olid up to everal inche in diameter, i good for low hear pumping, i elf-priming, ha a reverible flow direction, and ha only a ingle haft that require ealing (Figure 1 on next page). Progreive Cavity Pump Theory René Moineau, an aircraft deigner who wa trying to invent

2 Figure 1. A progreing cavity (PC) pump fitted for indutrial and municipal ervice Figure 2. Conventional (top) and long (bottom) geomtery ingle helix rotor and double helix tator deign an engine upercharger, deigned the PC pump in The range of Dr. Moineau patent i truly amazing. Many of hi deign have only jut recently become commercially available due to their manufacturing complexity. The mot familiar of the PC pump deign ue a ingle metal rotor, machined a an external helix, that revolve eccentrically within a rubber injection molded double internal helix that i twice the pitch length. The rotor i uually metal, the tator elatomeric with a compreion fit. However, where a tolerance fit i ued, rigid material uch a metal and platic can be ued for thee component. With rigid material, PC pump behave imilarly to other rotary PD pump uch a gear, crew or lobe pump. Thee deign are a pecial conideration and will not be covered in thi dicuion. Dicrete cavitie are created when the rotor and tator are combined. The cavitie piral or progre through the pumping element a the rotor turn. A one cavity diminihe, the following increae. The fluid cro ection i unchanged regardle of rotor poition, o it function like a piton in a cylinder of infinite length. One of the unique and defining propertie of the PC pump i that more than 90% of the loading i axial intead of radial becaue the liquid i moving along the ame axi a the rotating part (Figure 2). Lengthened Pitch Geometry Lengthened pitch geometry wa developed in early 1970 a a reult of the availability of whirling head lathe cutting

3 Figure 3. Double helix rotor and triple helix tator deign deliver 50% more flow but with 100% increae in internal velocitie machine in which the bar tock i held tationary and everal cutting tool are placed in a head that rotate or whirl around the bar tock. Becaue the cutting force were balanced it became poible to manufacture very thin and long rotor. The previou method of cam operated ingle point machine required that rotor be fairly hort and thick to minimize deflection caued by the cutting tool. In the new lengthened pitch geometry, cavity volume could remain unchanged while pitch length increaed and cavity and rotor diameter decreaed. Reduced rotor diameter and circumference reulted in reduced urface velocity at the ame rpm a the conventional geometry. The reduced diameter and rotor cro ection reulted in thrut load that were imilarly lowered. Thi i becaue thrut load (lb.f.) i a function of the cro ectional area of the rotor (in 2 ) time the differential preure (pi). The longer geometry alo increaed the ealing line cro ection while maintaining the ame compreion. Thi yield le lippage and higher volumetric efficiencie than the conventional geometry. Longer pitch geometry i ueful for abraive application becaue of it lowered velocity between the rotating and tationary component. It can alo accomplih low preure (< 60 pi) metering becaue it ha a flatter performance curve. It hould not be ued on high vicoity liquid, large olid or very low NPSHA application, however, becaue the entrance into the cavity i coniderably maller than on the conventional deign. Multiple Helix Geometry Dr. Moineau alo deigned the newet development in progreive cavity pump. Multiple helix rotor and tator became popular and have been ued for drilling mud motor ince about They became affordable for pump deign in 1993 when new whirling machine that could form double helix rotor became available (Figure 3).

4 Multiple helix deign can include an unlimited number of helice on the rotor and tator, a long a there i one more helix on one member than it mate. The mot affordable and practical of thee deign i a double helix rotor and a triple helix tator. In thi deign, the fluid cro-ection decreae by 25% but the internal velocitie double. Flow increae by 50% in the ame phyical pace. Additionally, in the 2/3 multiple helix geometry, tarting torque decreae with equivalent running torque. Thi deign i good for thin liquid and abraive, and it i excellent for variable frequency drive (VFD) application. It hould not, however, be ued on high vicoity liquid, to pump large olid or on very low NPSHA application. Thi i due to the reduced ize of the opening into the cavity in relation to the linear velocity of the liquid. Comparion to determine which pump i bet hould be between imilar pump tyle and model from different manufacturer. Internal linear velocitie and the velocity of the urface of the rotating part mut be compared, a well a the NPSHR, to determine which pump i betuited for a particular application. Certainly the new 2/3 geometry pump will provide the lowet cot per unit pumped, but they are everely retricted on other parameter. Mechanical Difficultie Unfortunately, PC pump are prone to a variety of mechanical difficultie. Thee can be claified into five major area: pumping element, univeral joint, haft eal, bearing and drive. Rotor Problem Eroion i the mot common problem for rotor. Operator hould try to ue the hardet material that i chemically compatible with the pumped liquid. Rotor coating, uch a chrome plating, ceramic and other pecial hard coating (which are harder then the rotor bae material) will all increae the longevity of the rotor. Unfortunately, ome care mut be taken with the coating. If the coating crack, peel or flake, the rotor will detroy the tator. Remember that the rotor i loaded with thrut by the dicharge preure, and the drive train of the pump i trying to keep the preure from expelling all of thee component out of the rear of the pump. The rotor bend a it rotate, ometime more than lightly, and brittle material can eaily crack. Sometime it i better to ue hardened bae material with no coating than to ue very hard but eaily fractured coating.

5 Another common way to increae rotor life i to derate the preure capability of the pump a the differential preure increae. While thi increae the initial cot of the pump, and decreae it mechanical and over-all efficiency, lowering the preureper-tage can dramatically increae rotor and tator life. Since the dicharge end of the tator i puhed away from the rotor by the preure of the liquid, the tator become conical on the dicharge ide of the pump. Thi reult in a reduced contact area on the uction ide of the rotor and tator. Uing more tage (or rotor and tator helix length) in the pump create more contact area between the rotor and the tator. It like adding tread depth on a tire; there i imply more material to wear away and part life i extended, ometime dramatically. While uing pump for higher preure can increae the Figure 4. Failure of chrome plating due to corroion of rotor bae material (carbon teel) purchae price by 20% or more and increae power conumption by 50% or more, thi election technique can extend the mean time between maintenance requirement by a factor of four to ix or even more. Corroion i alo a problem with rotor, and election of the correct bae metal i extremely important. Luckily, there are a variety of material compatibility guide available. If there i any quetion a to compatibility, the pump upplier hould be willing to provide tet coupon of any material ued in the pump, whether it i metallic, elatomeric, platic or ceramic. Coating failure, a previouly mentioned, are common. Hard chrome i ubject to poroity problem, and the cracking of ceramic i alo common. If the bae metal tart to corrode, the coating can eaily lift. It i critical that thee i no poibility of bae metal corroion. In many cae, where ome corroion of the bae metal i inevitable, rotor without any coating are the bet choice. In fact, very few Hatelloy or titanium rotor are old with coating becaue the fluid being pump will have ome corroive effect on the bae material over time (Figure 4). There are ome pecial rotor problem related to material failure. Generally, chrome plating

6 cannot be ued on hardened teel becaue of the hydrogen embrittlement of the bae metal during the chrome plating proce. The bae metal can crack and, unle X-ray inpected, the rotor can actually break in operation. Special coating can be ued with hardened rotor; however, ome of thee are proprietary. Corroion of hardened teel in intermittent ue can alo be a problem. Hardened teel are not good for application where the pump i routinely drained, a corroion rate are amplified with the increaed expoure to oxygen. It i alo important to ue eparate material in the u-joint to prevent galling. Thi i particularly true for deign uch a food grade pump, where the rotor i an integral part of the joint. Stator Problem Elatomer compatibility, along with dry running, i one of the mot common problem with PC pump. While the ue of an elatomer tator increae the utility of the pump for abraive or olid laden application, chemical compatibility can caue a whole new et of difficultie. Again, there are a variety of reference ource that can help with elatomer election. If there i any doubt a to the acceptability of a material, conduct an immerion tet that lat at leat two week. Make ure to conduct the tet at the temperature at which the pump will operate, a compatibility i dramatically affected by thi variable. In general, the pump can tolerate a 10% change in the hardne of the elatomer and a 5% change in the volume of the elatomer. Underizing the rotor diameter can eaily accommodate volume change due to temperature. However, a volume change from chemical aborption can not be imilarly tolerated. If the elatomer well due to chemical attack, it will uually continue to increae in ize, and it phyical trength will deteriorate, eventually cauing premature failure. Abraive reitance varie widely with elatomer. While Buna N (NBR) i the mot commonly ued and leat expenive bae polymer for tator, a relatively new compound, hydrogenated nitrile buna rubber (HNBR), i far uperior. Of coure, HNBR i coniderably more expenive than NBR. Jut becaue a material i chemically compatible, there i no guarantee that the material will provide the longet life. Fluoropolymer (FPM) are notoriouly poor for the phyical propertie needed by PC pump. There are compound that are both le expenive (HNBR) and

7 more expenive (Afla ) that might be better choice. The choice ultimately depend on the material pumped. Certain compound, buna among them, and epecially white compound that are filled with kaolin clay a oppoed to carbon black, are hydrophilic. Pure water, deionized and ditilled water application require pecial elatomer. Another problem with ome elatomer i the high mineral oil content (Thi i one reaon why old truck tire burn o well.) Some hydrocarbon can break down thee oil over time and caue the elatomer to hrink and harden. A final concern of tator compatibility i the preence of gae in olution in the pumped liquid. Some gae, like CO 2, can be aborbed by the tator in their compreed or liquid tate. Thi ha no effect on the performance of the pump or the elatomer, but if preure i relieved from the ytem and the material change phae from liquid to ga, it will completely detroy the elatomer. Bonding adheive failure are alo common problem, epecially with tator that are cut-toize. Some manufacturer injection mold their tator in long, multiple-tage tick and then cut off the number of tage needed. Thi expoe the adheive that hold the elatomer inide the metal tube to the material being pumped. Thi may not be a problem for many application. However, it i a common problem with EPDM elatomer on application with ketone. EPDM i reitant to a wide range of chemical, including many adheive, and mot adheive are eaily diolved by ketone. Separating the adheive from the Figure 5. Rubber to metal bond failure on a cut-to-ize rotor pumped liquid by molding an integral rubber lip at the end of the tator i the bet way to prevent adheive failure. Gaket tend to not be very effective becaue the rubber i uually pulled away from the metal when the tator i cut. Thi can alo be a problem if carbon teel tator tube are ued for application where all of the wetted part are made of tainle teel or other corroion-reitant metal. Bonding adheive failure are alo common on jacketed pump, a heat reduce the trength of the

8 adheive holding the rubber inide the metal tube (Figure 5). Controlling the temperature at a point below the manufacturer recommendation for the elatomer i important. Thermotat for hot oil, water or reitance heater, a well a preure regulating valve for team jacketed pump, are an abolute requirement when heating the pump, a heat detroy the bonding chemical. Another tator failure problem i named after the hyterei vicoity curve ued to meaure the cure rate of the elatomer during vulcanization. Exceive flexing of the rubber due to high temperature and/or preure, which activate the vulcanizing enhancing additive in the elatomer and can caue additional hardening, caue hyterei failure. An eaier undertood explanation i the exploding truck tire phenomenon. No one ever ee big chuck of truck tire inide town. They are alway on the highway. If a truck i overloaded, with too much compreion on the elatomer tire, it can run at low peed without a big problem. If thi ame truck run at higher peed on the highway, the elatomer i compreed and relieved at a higher frequency. Heat build up that can caue the elatomer to further cure and become hard, like platic. Without reiliency the elatomer tart to break apart and big chunk of rubber come off the tire, jut like they will come out of the end of the pump. To avoid thi problem, the rotor mut alway be ized for the correct temperature and the pump hould never be expoed to higher than rated preure. Hyterei failure can be caued by repeated dead head operation that take place for hort period of time but at high frequencie, which i why it i alway important to cycle valve when the pump i not operating. Run Dry Damage The mot common problem with tator i probably run dry operation. The compreion fit ued with elatomer tator need lubrication to carry away heat. Heat will build up if there i no fluid, or if there i fluid but it i not moving through the pump. In thi ituation, it i heating up to above the temperature limit of the elatomer. There are variou device available to prevent the problem, but all of them have limitation. Thermitor can be imbedded in the tator to meaure the operating temperature, and a et point controller can be ued to hut off the pump at a pecific tempera-

9 ture. Thee device are inherently very reliable, but are only available a part of a new tator; they cannot be ued with adjutable tator; they are not allowed in dairy (3A) application, and they cannot be ued on very mall PC metering pump (Figure 6 on next page). Paddle-type flow witche can be ued, and they are very inexpenive. However, they are not recommended on vicou or olid-laden material, which are a common application area for progreive cavity pump. Heat tranfer flow witche are reliable on a great number of application, but they are generally more expenive than the thermitor device and are not uitable for mot exploion proof application. There are a number of fluid detector that can be ued to protect the pump. Keep in mind Figure 6. Thermitor imbedded in tator to protect againt run-dry damage that thee device generally only meaure the preence or abence of a fluid and not heat, which i the real caue of tator failure. Thee device hould alway be intalled with a preure witch to prevent either dead head or cloed uction condition. They alo have to be intalled in a vertical or elf-draining line to be effective. Capacitance-type detector are about the ame price a a thermitor ($650), and are available in a food grade deign at a omewhat higher price. Some cutomer have ued tuning fork fluid enor, but they need to be inpected periodically becaue they can foul and become unreliable. They are, however, le expenive than the capacitance type fluid detector. It i alo poible to integrate flowmeter into proce intrumentation to prevent run dry operation. Thi i the leat expen-

10 ive way to protect the pump, provided you can get the equipment intalled, programmed and operating before tart-up. Preure witche can alo be ued, with ome precaution, to protect the pump from run dry damage. A witch, ued on the uction ide with a minimum preure et point, can prevent damage to both tarved uction and emptied uction tank (if the tatic fluid level i high enough to detect thee condition). Many cutomer alo like to draw down tank to empty, but there i not enough enitivity in a tandard preure witch to draw a tank down to zero tatic head without running the pump dry. A preure witch, rupture dic or preure relief valve hould alway be ued on the dicharge ide of the pump to protect from pumping againt a dead head. Thee pump are poitive diplacement pump, and if they are operated againt a dead head, they will try to build preure until one of the component or the piping fail. Preure witche can protect againt running the pump dry by uing a dual et point witch that protect againt both over and under preure condition. Dicharge preure i a combination of both tatic head and friction lo. If a uction line i cloed or the feed ource emptie, the friction component of the total head will diappear and the recorded preure will drop. Thi ytem i not very reliable when the friction lo component of the total head i very low. Univeral Joint Problem While ome manufacturer try to make univeral joint deign a major difference between progreive cavity and other pump, they generally do it to give themelve an advantage in written pecification. The truth i that for mot application, it i only important to enure that the univeral joint are poitively ealed from the pumpage and are properly lubricated. Sealed u-joint are imperative, a any auto mechanic working on front wheel drive car can tell you. There i alo no real ue in having a univeral joint that will not need rebuilding before the rotor need replacing, ince you have to diaemble the joint to replace the rotor. Generally, it i alo adviable to have acrificial or low cot wearing part in the joint. Thi will minimize the cot of repair and make it fairly eay to work on. Gear type u-joint are notoriou for their high replacement cot and repair difficulty, but they are very good for high thrut load

11 Figure 7. Gear type and ealed pin-type univeral joint in comparable ize application. Remembering that thrut load i a function of preure and rotor diameter, high thrut load are only een in very high preure ituation (> 1000 pi) or in very large pump (rotor Ø > 6 ). Cardan joint can alo be ued in thee application, but they have to be oil filled, the ame i true for ome gear type univeral joint. For 95% of PC pump application, imple ealed and greae lubricated pin joint, which are made by almot all of the major PC pump manufacturer, are acceptable (Figure 7). Exceive thrut load caued by high differential preure ( P), a mentioned above, can caue problem with the univeral joint. On abraive application the tator and rotor will normally fail before the univeral joint. The oppoite i true for non-abraive, ambient or lower temperature application the joint will fail firt. The weaket component in thee cae i uually the joint lubricant. The high thrut load combined with friction between the joint component can produce enough heat to vaporize the lubricant, which i why ome manufacturer in their larger or higher preure pump ue oilfilled joint. It i alway wie to ue only the manufacturer recommended lubricant in the joint and check the operating preure, epecially on the econd and third hift operation where plant peronnel are alway more willing to ue throttling valve than variable peed drive to control flow. Suction preure can caue joint problem a well. If the eal that protect the joint from the pumpage i damaged or diplaced, the joint will fail. Many gear joint are limited to 25 pi in the uction caing. There are ome double eal deign that can handle up to 50 pi. Hydraulically balanced ealed pin joint are available that can operate up to 175 pi. Thi i omething that you really need to conider if you are operating the pump in revere, uch a with a uction lift application or when

12 pulling againt a vacuum, i.e., a crytallizer or concentration unit. The dicharge preure will affect the univeral joint. Univeral joint angularity, again, i a point puhed by ome manufacturer to try to exclude competition. The truth i that the correct angularity varie with the type of univeral joint. The angularity mut match the u-joint deign. For gear joint it hould be < 2, becaue they generate a lot of heat due to the thrut plate, which enable them to aborb the higher thrut load. Pin joint hould have an angularity of around 3. Cardan joint need to have an angularity of > 5 to enure that the needle bearing inide of each bearing cup rotate to promote even wear. Shattered joint part are a ign of cavitation, and thi happen in any type of univeral joint in a PC pump. The imploion of vapor bubble on the dicharge of the pump caue high momentary thrut load on the drive train. Mot manufacturer ue ome hardened teel or cat iron component in their joint, and the hock load caued by cavitation can fracture thee part. In gear joint with oft alloy thrut plate, galling and deformation leading to failure of the joint eal i common with cavitation. Pump with flexible haft, intead of u- joint, will break thee haft. Cavitation hould be avoided in any pump, and while Dr. Moineau originally deigned PC pump a compreor, cavitation will have a damaging effect over time. PC pump can withtand ome cavitation becaue they operate at low peed. There i a preure gradient a the fluid travel the entire length of the pumping element, and the drive train i mounted in a reilient mounting (the elatomer tator). Cavitation, however, will have a negative effect on your pump. Sealing Problem PC pump are famou for not having ealing problem. Compared to many other rotary PD pump, they have an advantage becaue they only have one haft to be ealed. Thi eal i uually only expoed to the uction preure condition and 95%+ of the total load i axial rather than radial. In cae with low NPSHA or uction lift application, a mentioned previouly, it i adviable to run the pump in revere. Thi prevent air from being ucked in through the packing or the mechanical eal. A fluh or a quench can help alleviate thi problem. Failure to prevent thi lo of prime through the ealing

13 area can lead to run dry damage of the tator. When operating with packing, it i important to fluh abraive with a lantern ring pacer in the packing and an API plan 32 or 52 fluh. Thee fluhing ytem create a preure barrier in the packing that prevent abraive from entering. They alo keep the packing cool and lubricated, and thi prevent damage to the pump haft. There i a myriad of packing available, and they can be fitted to any PC pump. Your PC pump manufacturer or local ditributor can intall your favorite packing for you. Like pump vendor, if you have a packing upplier that you trut and provide good ervice, ue hi or her recommendation. Your truted pump upplier will repond in kind. For packing, though, it i alway recommended that you ue a hard coated haft or haft leeve. Packing rub againt the haft and where there rubbing, there wear. The harder the haft, the longer it will lat on packing. Single eal are gaining in popularity. For eepex, more than half of all the pump we ell have ingle eal. The price of hard faced, ilicon carbide or tungten carbide eal, epecially rubber bellow eal, have gone down dramatically over the lat everal year. Generally, the price for a ingle eal i equal to the cot of a good packing with a hard coated drive haft and a lantern ring eal fluh. In ome PC pump deign, it i actually eaier to replace a eal than it i to replace packing. There are a few caveat with ingle eal. You hould ue a quench if there i more than five foot uction lift when running the pump in tandard rotation. A quench i jut an area behind the eal face, ealed by a rubber lip eal, where a clean liquid (water or machine oil, generally) i preent at atmopheric preure. Thi form a vicou barrier that prevent air from entering through the eal when the uction preure in the pump i at le than atmopheric preure. The quench can alo be ued to prevent crytallization of the product on the eal face. Single eal with quench have proven to be effective on paper coating, paint, ugar olution, honey, alt olution and a variety of other crytallizing material that are known for detroying packing and ingle eal without a quench. On certain application it i important to ue lurry eal houing for abraive. It i adviable to keep thee hard faced eal cool by placing the eal inide the uc-

14 tion caing of the pump. Ceramic or carbide eal face can fracture or chip if they are allowed to get too hot. The pumpage keep the eal cool and prevent the build up of olid around the eal. Thi cannot be achieved by placing the eal inide a tuffing box deigned for packing. Some PC pump can eaily accommodate thi eal mounting arrangement. Other, which have the tuffing box cat a part of the uction caing, require expenive modification to accept thi eal. Double eal, of coure, are till widely ued. However, mot double eal in ue now come in the form of a cartridge eal. Double eal are till widely ued on fine abraive (like pigment) and crytallizing material (like latex), which can foul ingle eal with a quench. Double eal mut have a fluh ytem with the eal fluh liquid at a preure that i uually 5 pi higher than the preure in the uction houing. Again, an API plan 52 fluh ytem, which can cot a much a a mall PC pump, need to be intalled with a double eal. If a double eal i not fluhed, it can be detroyed within minute. One lat word regarding double eal: Split cartridge eal may not fit into ome PC pump. Pleae be careful when uing thee eal, which have very large diameter gland. Some manufacturer are making unit that will fit. Other may fit but may require modification to the uction caing or the bearing houing of the pump. The added cot may not be worth the added convenience of the plit eal. Bearing Problem Bearing problem, becaue o much of the load i axial rather than radial, are not a evere or common problem in PC pump. Firt, new bearing alway run hot (up to 160 F), and they will take everal day to run in. Becaue of the high axial load, avoid ball bearing except on very mall pump. Tapered roller bearing can handle higher thrut load and are a better invetment. Proper fitting of the bearing cover plate i important, and while bolted plate are eaier to ue than nap ring fitted plate with adjutment him, the tolerance are the ame. Once properly et and lubricated, you can expect tapered roller bearing to lat a long time. Cloe coupled PC pump are becoming more popular. In Europe, the majority of PC pump old are cloe coupled or block configuration. Thee arrangement have made it poible for PC pump upplier to reduce the cot to uer by a

15 Figure 8. Integral or block pump contruction (top) where gear reducer bearing aborb pump thrut and radial load much a 40%. It ha alo helped to reduce lead time. Additionally, it i afer than V belt and pulley, and more reliable. It a great idea. But like a lot of great idea, it can be abued. Thee unit are typically oil lubricated. If a manufacturer propoe a cloe coupled pump, make ure that the gear box i rated with, at the minimum, a 1.5 ervice factor baed upon the motor input power. Thi i the minimum factor according to AGMA for cla II gearing with a poitive diplacement pump. Secondly, enure that the pump manufacturer provide you with the actual maximum thrut load calculation and the maximum thrut load rating of the gear box. The gear box hould be rated for more load than the actual calculated load. Otherwie, ue a larger gear box or a pump with dedicated bearing (Figure 8). Protection of the bearing from contamination i alway a major concern. Ue IP65 double lip eal protection or labyrinth eal for the bearing on fine lurry or coating type fluid to prevent contamination. Becaue of the extremely low peed ued with PC pump on highly abraive application, haft linger ring can be uele. High quality fluid-end and bearing eal will olve mot bearing problem aociated with PC pump. Drive Problem Until about a decade ago, variable pitch pulley belt drive were the mot popular drive for PC pump. They are till one of the leat expenive way to achieve variable flow in a rotary pump. Unfortunately, there are ome izing and reliability concern with thee drive. It i very important to enure

16 that the proper hp rating i ued. Some manufacturer rate thee unit on their input power, and ome rate them on the output power. Thi can be a big difference, due to the mechanical loe aociated with both the variable pitch belt and the gear reduction unit. Make ure that you have enough power at the drive output to power the pump. Since thee unit are invariably integral with a gear reduction unit, make ure that the ervice factor of the gear box at low rpm i ufficient. Becaue they are a mechanical drive, the torque increae a the peed i reduced. Some unit may have very low ervice factor on the gear at low peed ome may be le than 1.0. It i recommended that cutomer buy the hardened or hard coating pulley option on thee drive. The belt ha a tendency to run a groove in conical pulley. Mot manufacturer recommend that uer run the drive all the way to it maximum peed and all the way down to it minimum peed once a week to prevent grooving. I ve never known a uer to have thi tep in their formal PM procedure, o the hard coating i a worthwhile afety precaution. One of the mot common problem with gear boxe and mechanical drive, which are integral with gear boxe, i venting. If the box i not vented, vapor will build up and the preure will blow out the oil eal. Be ure gear box breather, which are hipped eparately to prevent lo of lubricant during hipping and intallation, are intalled. Failure to intall breather i probably the mot common caue of gear box failure. Mechanical friction drive can be ued on PC pump, but it i important to enure that the drive i rated for maximum pump tarting torque and not the maximum running torque. The hock load aociated with PC pump can hatter the phenolic friction ring in thee drive if they are underized. Again, be ure that the gear box breather are intalled. Electronic drive have become extremely popular for all variable peed application in the lat decade, including PC pump. While the electronic have eliminated a lot of the problem aociated with mechanical component, they have given rie to other problem. Progreive cavity pump are contant torque device, if the differential preure i contant. Therefore, if a variable frequency drive (VFD) i ued, it mut be of the contant torque type. VFD mut be ized conider-

17 ing the drive tarting torque capability a well a it operating torque capability. VFD cannot generate a much tarting torque a a mechanical variable peed drive, and PC pump, due to the compreion fit between the rotor and the tator, require a lot of tarting torque. Depending on the operating preure, tarting torque i uually 50% higher than the running torque and can be a high a two or three time the running torque if the pump preure capability ha been everely derated to improve rotor and tator life. Ue of the following formula to calculate the proper drive ize, given the pump maximum peed and tarting torque in lb. ft.: (lb. ft. x rpm)/(5250 x tarting torque current boot) = VFD horepower VFD are baically computer, and they can be difficult to program. Of coure, each VFD i different; but there are a few guideline to remember: Set the current boot for tarting to the maximum etting. Minimize the ramp up and oft tart capabilitie. Thi increae the amount of tarting torque. Locate the VFD a cloe a poible to the motor. Problem tart to arie when the VFD i more than 100 feet from the motor. For more turn down (>6:1), et the maximum pump 90 Hz with four- or ix-pole motor. Thi will enable you to ue a higher reduction on the gear box, which provide more torque for tarting. It will alo have the motor running at higher peed to allow for improved cooling of the motor. Jut about any type of prime mover including, but not limited to, air motor, hydraulic motor, DC motor, gaoline and dieel engine can drive PC pump. There have even been hand-operated PC pump. Remember to match both the tarting and running torque requirement, and ize the unit over it entire operating peed range. Application Problem Some very pecific application problem need to be mentioned to finalize thi topic. High vicoity application uually require open hopper pump. It i important to ize the hopper o that material will not bridge in the unit. Thee pump are available in a variety of deign depending on the particular application condition. They are available with and without extenion tube induction zone to improve the volumetric efficiency of the pump, and they are available with

18 Figure 9. Open hopper type pump ued for high vicoity liquid and ludge. Note auger feed device. internal cutter for chopping up potatoe or beet or fruit. Special auger, ome with mixing capabilitie, can be intalled a part of the drive train, and eparately driven bridge breaker can be intalled in the uction caing of the pump. Thee not only protect againt bridging of the product above the open hopper pump auger, but will impart additional hear to lower the apparent vicoity of thixotropic or peudo-platic fluid. Thi make them eaier to pump (Figure 9). Exceive peed i another common caue of premature PC pump failure. Again, rpm i not a ueful meaure of peed. What i important i the urface velocity of the rotor againt the tator. Necearily, a higher flow are needed, the cavity in the pump increae in both length and diameter, and the rotor ummarily increae in diameter and circumference. Some broad guideline for maximum peed relative to the flow rate required are: > 500 gpm = < 250 rpm > 50 gpm = < 350 rpm > 5 gpm = < 500 rpm > 0.5 gpm = < 1000 rpm Cavitation, a mentioned previouly, can caue problem with the univeral joint, and it can damage the rotor and tator. Don t ever intall a pump without comparing NPSHA to NPSHR. Even though the uction may be flooded, high vicoity application commonly have enough friction lo aociated with valve and piping to reduce the NPSHA to a level below the NPSHR. It i alo important to know the tandard being ued by the pump manufacturer to meaure NPSHR. There are difference between the tandard ued by the Hydraulic Intitute, API and other organization. The pump may not produce according to the publihed performance curve, even if the NPSHA i above the manufacturer lited NPSHR, becaue of the tandard ued to meaure NPSHR.

19 It i important to intall gauge on both the uction and dicharge ide of the pump. Some manufacturer do not have gauge connection on their pump or they may only offer them at an extra charge. It i impoible to diagnoe a pump problem in the field without preure gauge. Preure i, after all, one of only two component that define work in a pump. If no proviion i made for gauge intallation, your piping will have to be changed. It like flying in a now torm without an altimeter. Exceive preure and/or temperature in an application will caue everal of the earlier lited condition: tator hyterei, u- joint or bearing failure. It i imperative that you know the temperature and preure for your PC pump intallation. One of the thing that PC pump are not good for are application with wide temperature fluctuation. Thi i due to the expanion and contraction of the elatomer. High temperature will caue either increaed eroion rate or hyterei failure of the elatomer. Low temperature will caue reduction of the compreion in the pumping element, exceive lip and neceitate premature replacement of the tator and/or rotor. Proper election of the elatomer material will help to minimize thi problem, but application with temperature fluctuation of more than 150 F are generally not recommended for PC pump. While PC pump manufacturer like to promote low peed to increae the life of the pumping element, thi can backfire on ome heavy and hard-olid laden lurrie where there i inufficient internal velocity, and the olid ettle in the pump cavitie. In thee intance, rotor will wear a fat or fater than tator, u-joint cover can get hole worn in them and fail, and tator will quickly erode. The bet olution i to intall an auger or propeller type coupling rod to add turbulence and prevent ettling. Uing a long or multiple helix geometry will alo increae the linear velocity in the pump while till keeping the urface velocity of the rotor low. Concluion Progreive cavity pump are veratile and adaptable for a wide range of application. Unfortunately, they are omewhat more enitive than other more commonly ued pump. If care i taken with proper election and intallation, PC pump can be a uperior choice. Chooing the proper material of contruction, peed, geometry, eal and drive are only part of the job for a good

20 Return pump intallation. Proper intallation require the incluion of a reliable and appropriate device to prevent dry running and offer preure protection. In addition, proce temperature control and properly programmed electronic drive will enure that your PC pump i dependable for a long time. Michael L. Dillon i Vice Preident and General Manager of eepex, Inc., a manufacturer of progreive cavity pump in Enon, Ohio. Mr. Dillon i the author of numerou article on poitive diplacement and progreing cavity pump. Thi article i baed on a paper preented at PumpUer Exp 98, ponored by Pump & Sytem Magazine. REPRINTED FROM PUMPS AND SYSTEMS MAGAZINE