 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Pet Forum / Aquaria / Marine Reef / March 2004Throttle a pump harmful?
I've read here and on Marc's tank website that you shouldn't throttle centrifugal pumps with a ball valve but that you should divert some of the flow back to the intake side. I hope I'm not being obtuse when I ask, "Why?" Speaking specifically about Mag pumps, the manufacturer doesn't warn about or otherwise recommend any minimum flow rates through the pump, and the backpressure from a constriction in the line should be identical to the backpressure from a higher head application. At least that's what my math comes out to. I don't think the pump can tell if the reason it's not pumping as much is because there's a 5' head or an almost closed gate valve in the line. If I'm wrong about this, could someone please take the time to explain how the two cases (partially closed valve vs. increased head) are different from the pump's perspective? Or, alternatively, something else (not necessarily pump damage) may be the reason to keep the flow through rate at a maximum. Perhaps something like plankton mortality rates (just stabbing in the dark here) that get much worse if the backpressure goes beyond a certain point... Regards, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller Ross, my whole reasoning is to avoid putting any undue stress on the pump, and that is my only reason. Why add head pressure to a pump that doesn't need it? I believe that the extra effort may lead to the pump running at a higher temperature, possibly warming the tank water more. Marc > I've read here and on Marc's tank website that you shouldn't throttle > centrifugal pumps with a ball valve but that you should divert some of [quoted text clipped - 27 lines] > "Security is mostly a superstition. It does not exist in nature... > Life is either a daring adventure or nothing." -- Helen Keller -- Personal Page: http://www.sparklingfloorservice.com/oanda/index.html Business Page: http://www.sparklingfloorservice.com Marine Hobbyist: http://www.melevsreef.com Ross / Marc Do you have a pump curve? It's a long time since I was doing chem. eng. but my memory says the pump curve gives flow / head / efficiency and power consumption. Until you have a look at that you don't really know which condition will generate more heat. I think stress on the pump should not be an issue -- they should be designed to run within their stated range. Phil > Ross, my whole reasoning is to avoid putting any undue stress on the pump, and > that is my only reason. Why add head pressure to a pump that doesn't need it? [quoted text clipped - 39 lines] > Business Page: http://www.sparklingfloorservice.com > Marine Hobbyist: http://www.melevsreef.com Hi Ross For most impeller pumps and centrifugal pumps it wouldn't matter at all, because they are designed to operate within a wide range. But there is a better way than clamping down the output feed line. Install a T-Fitting in the output line and a line connected to the T-Fitting as a return line to your sump, you can install a valve or clamp this line to increase output from the feed line. This method works well on all pumps, keeps heat buildup lower and places less stress on the pump. We have similar set-ups on all of our bottle filling equipment, except instead of a manual valve it has a spring loaded ball valve that can be set at various pressures. When the filling head solenoid opens, you have the desired head pressure. When the filling head valve closes, the spring loaded valve is forced open with the excess head pressure and allows the product to recycle back to its own carboy (sump). Some systems use a split solenoid so that when one side is open the other side is closed, but you get unequal head pressure for a split second as the solenoid switches, which can cause a splash of the product, so these split solenoids are rarely used. In fact, most small bottlers and repackagers use gravity feed rather than pump feed to save costs on electric and equipment replacement costs. TTUL Gary [...snip...] > But there is a better way than clamping down the output feed line. > Install a T-Fitting in the output line and a line connected to the > T-Fitting as a return line to your sump, you can install a valve or > clamp this line to increase output from the feed line. > This method works well on all pumps, keeps heat buildup lower and > places less stress on the pump. Thanks for the answer. This does respond to the core of the question that I was asking. So what you're saying is that operating a pump at a higher head does a few things: 1) increases wear/stress on the impeller/motor 2) increases heat production/reduces efficiency Both sound reasonable and plausible, but I have heard that lower flow rates can make some pumps work less (that they can work more efficiently at heads greater than 0ft than they do at 0ft). This has been asserted for the Rainbow Lifegard Quiet One pump on this very newsgroup. This assertion is also plausible if the efficiency of a pump is nonlinear (goes up at lower pressures, then drops again at higher pressures, going back to zero at the pump's max head). Now, what I really wonder is: does anyone have any actual numbers to support either set of assertions. These numbers might only apply to a particular make/model of pump, but any empirically gathered numbers would help to satisfy my curiousity. Regards, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller Ross You need the pump curve for the particular pump to answer those questions. Phil > [...snip...] > [quoted text clipped - 31 lines] > "Security is mostly a superstition. It does not exist in nature... > Life is either a daring adventure or nothing." -- Helen Keller > Ross > > You need the pump curve for the particular pump to answer those questions. Given these pump curves, what can you tell me? Or is there another curve (more/different data) that would be needed to answer my question? http://www.marinedepot.com/aquarium_pumps_pentair_aquatics_rainbow_lifegard_quie t_one_information.asp#qone800 There's definitely a "belly" to the curves and it seems that if you chose a point where the area within the rectangle of the flow and height is maximized, you may have found some sort of a sweet spot. But is that sweet spot likely to have the lowest wear on the pump? Is the pump likely to run most efficiently and produce the least waste heat at that point on the curve? What might that sweet spot mean to us as aquarists? Thanks for all the help, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller > Given these pump curves, what can you tell me? for what you are after, no one can tell you anything from those graphs > Or is there another curve > (more/different data) that would be needed to answer my question? very much so, you need one that has that curve and some kind of electrical consumption curve(you could get the answer from almost any electrical curve), it would be nice to have a MTBF curve added to that, another might be heat transfer. when water moves slower thru a hot object it picks up more heat some of that is set, some of it depends on what the water is moving thru. > There's definitely a "belly" to the curves and it seems that if you > chose a point where the area within the rectangle of the flow and > height is maximized, you may have found some sort of a sweet spot. > But is that sweet spot likely to have the lowest wear on the pump? Is > the pump likely to run most efficiently and produce the least waste > heat at that point on the curve? What might that sweet spot mean > to us as aquarists? that sweet spot probibly doesnt mean anything, except where the most gph is. most aquarium pumps dont list the data you want. some larger like 1/4+ hp pumps do.  SignatureRichard Reynolds Richard.Reynolds@usa.net > > Given these pump curves, what can you tell me? > > for what you are after, no one can tell you anything from those graphs That's kinda what I figured. I'm getting the distinct impression that I would have to buy several different pumps and do some testing myself to get real answers to my questions. Possible test chassis configuration: 1) A temperature controlled sump. 2) Demand flow from the sump into a test chamber with the pump, or a line directly from the sump to the pump input through a bulkhead (depending on submerged/external). 3) A temperature sensor on the side of the pump casing. 4) A pressure sensor at the pump output connection. 5) An ammeter on the pump's power supply. 6) An output tank incorporating a simple flow meter. 7) A rack so the output tank can be moved up and down (this will be impractical in my garage for heads over 8' but that may be enough). 8) A temperature sensor in the output tank. 9) Return line to the temperature control systems of the sump. 10) A data-aquisition system to record all of this data (so I don't have to do manual data recording). The experiment variables would be the pump model selected, the temperature of the sump, and the height of the head. The experiment could then be varied to include reduced pump flow strategies (like the throttle valve vs. shunt approaches which are the topic of this thread). > it would be nice to have > a MTBF curve added to that, another might be heat transfer. Well, the MTBF curve will have to be sourced from the manufacturers, even if I do conduct this experiment. When I worked at Texas Instruments in the early 90's I helped them summarize the device qualification data to determine MTBF and there's simply no way I could afford to buy and run enough pumps to get significant accuracy on an MTBF number (hundreds of test units minimum). What's really too bad is that MTBF doesn't appear to be available for any of the smaller powerheads that are so popular among aquarists. Now to see how much this experiment would cost... Regards, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller > That's kinda what I figured. I'm getting the distinct impression that > I would have to buy several different pumps and do some testing myself > to get real answers to my questions. maybe. > Possible test chassis configuration: > 1) A temperature controlled sump. maybe but probibly not. > 2) Demand flow from the sump into a test chamber with the pump, > or a line directly from the sump to the pump input through a > bulkhead (depending on submerged/external). ok ish > 3) A temperature sensor on the side of the pump casing. but if your gona do it you have to have 2 one at the input one at the output and compare the difference, just because heat is generated doesnt mean its xfered to water. though a 3rd on the casing might be interesting it isnt the heat you need. > 4) A pressure sensor at the pump output connection. and a gate valve youd wanna adjust to simulate a different head. > 5) An ammeter on the pump's power supply. yep > 6) An output tank incorporating a simple flow meter. sounds too complex to me, just attach flow meter on output end of gate valve. > 7) A rack so the output tank can be moved up and down (this will be > impractical in my garage for heads over 8' but that may be enough). nah to impractical head height and PSI are different measurements of the same thing you can simulate "head" by increasing the PSI load on the pump using a gate valve. > 8) A temperature sensor in the output tank. to far from the pump > 9) Return line to the temperature control systems of the sump. that part really doesnt work in my brain, but dont have a good reason why yet. > 10) A data-aquisition system to record all of this data (so I > don't have to do manual data recording). YEP!!!! > The experiment variables would be the pump model selected, the > temperature of the sump, and the height of the head. > > The experiment could then be varied to include reduced pump flow > strategies (like the throttle valve vs. shunt approaches which are the > topic of this thread). :D > > it would be nice to have > > a MTBF curve added to that, another might be heat transfer. [quoted text clipped - 5 lines] > afford to buy and run enough pumps to get significant accuracy on an > MTBF number (hundreds of test units minimum). more like thousands you would have to pick common head heights and test hundreds of each, common are like 0, 1, 4 ... but yea. > What's really too bad is that MTBF doesn't appear to be available for > any of the smaller powerheads that are so popular among aquarists. I know you can get them for rio, and mag drives, you can actually find some of this already for different pumps in different conifgurations. > Now to see how much this experiment would cost... $$$$$$$$$ SignatureRichard Reynolds Richard.Reynolds@usa.net Ross These curves are simple and don't tell you much. You also need efficiency and power consumption data. For some general info, have a look at http://www.mcnallyinstitute.com/06-html/6-01.html In my experience, centrifugal pumps are generally designed to be throttled by valves, and won't wear out because of it -- certainly energy loss across the valve restriction will generate some heat. Phil > > Ross > > > > You need the pump curve for the particular pump to answer those questions. > > Given these pump curves, what can you tell me? Or is there another curve > (more/different data) that would be needed to answer my question? http://www.marinedepot.com/aquarium_pumps_pentair_aquatics_rainbow_lifegard_ quiet_one_information.asp#qone800 > There's definitely a "belly" to the curves and it seems that if you > chose a point where the area within the rectangle of the flow and [quoted text clipped - 11 lines] > "Security is mostly a superstition. It does not exist in nature... > Life is either a daring adventure or nothing." -- Helen Keller > Ross > These curves are simple and don't tell you much. You also need efficiency > and power consumption data. For some general info, have a look at > > http://www.mcnallyinstitute.com/06-html/6-01.html That's an interesting read. Thanks for the link. > In my experience, centrifugal pumps are generally designed to be throttled > by valves, and won't wear out because of it -- certainly energy loss across > the valve restriction will generate some heat. To be sure, but is the heat generated an inevitable "feature" of the selected pump, and if not, does either throttling or shunting minimize the generated heat? In another response, I have proposed an experimental harness to test the question, though I won't be able to test for increased wear and tear any more precisely than "signs of wear" (damaged impellers, etc.). I'm actually thinking about doing this experiment. I am a crazy man... Regards, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller Ross, I replied to your email, but it came back "delivery failure" so I'll just post my comments here: As you know, I'm just a hobbyist and go on reasoning more than anything usually. I do keep an open mind and am willing to adjust my own systems if the advice turns out ot be different than my own configuration. Please do let me know what you end up finding out. So I understand your plan, do you plan to put a pump in a bucket and another bucket up at 5' (for example) and pump it full throttle and then try it again with a ball valve restricting flow? Keep in mind that if you do the Tee system as I recommend, you need to isolate that so the return pump isn't pulling in air bubbles. Matter of fact, you'll need to isolate the pump anyway to avoid airbubbles from the water draining down. Then you plan to test water temperature? Better keep the ambient room temp. the same if possible so that the test takes place under the same conditions. Marc > Marc Levenson <melev@swbell.net> writes: > [quoted text clipped - 19 lines] > Regards, > Ross -- Personal Page: http://www.sparklingfloorservice.com/oanda/index.html Business Page: http://www.sparklingfloorservice.com Marine Hobbyist: http://www.melevsreef.com > Ross, I replied to your email, but it came back "delivery failure" > so I'll just post my comments here: Hmmm. That's frustrating. ross@rossbagley.com should work. I just sent a test message on a quick loop no problems... > As you know, I'm just a hobbyist and go on reasoning more than > anything usually. I do keep an open mind and am willing to adjust > my own systems if the advice turns out ot be different than my own > configuration. That's the attitude that I have as well. Though it's often easier to plan ahead than to adjust later. On a new project that I'll be starting next year, I'm trying to plan, plan, plan as far ahead as is practical. Understanding how the sump plumbing will go together is the current topic, which is why I started this thread. > Please do let me know what you end up finding out. So I understand > your plan, do you plan to put a pump in a bucket and another bucket > up at 5' (for example) and pump it full throttle and then try it > again with a ball valve restricting flow? A little more sophisticated than that, but that's the initial plan. Basically, the goals will be to measure pump efficiency, water heating and impeller wear (qualititive only) for different pumps at different moderate heads. Then I intend to modify the experiment to include various means of reducing pump flow and obtain equivalent data under those circumstances. > Keep in mind that if you do the Tee system as I recommend, you need > to isolate that so the return pump isn't pulling in air bubbles. > Matter of fact, you'll need to isolate the pump anyway to avoid > airbubbles from the water draining down. This was the plan, but it's good to be reminded that the return circuit could introduce more bubbles. > Then you plan to test water temperature? Better keep the ambient > room temp. the same if possible so that the test takes place under > the same conditions. The ambient temperature will be recorded and controlled as much as practical. Water temperatures will be controlled within very tight parameters (and will be treated as one of the experimental variables). Living in SoCal without A/C means that I'm going to need a decent chiller for the next big project so if I get approval, I'm just going to go ahead and get it now to use it for the experiment. Regards, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller > > Ross > > These curves are simple and don't tell you much. You also need efficiency [quoted text clipped - 18 lines] > > I'm actually thinking about doing this experiment. I am a crazy man... YEP -- SOUNDS THAT WAY!!!!!!!!!! Phil > Regards, > Ross > > -- Ross Bagley http://rossbagley.com/rba > "Security is mostly a superstition. It does not exist in nature... > Life is either a daring adventure or nothing." -- Helen Keller Hi Ross It's very simple to hook up an ammeter to check the current draw as well as a thermometer to check motor temperature. It only makes sense, that to do MORE work, will require MORE energy and produce MORE heat. We use MaxiJet1000's to pump heavy viscous liquid, they are the coolest running of all submersibles we have tried, but do run right at their automatic shut-off pre-set temperature when the backpressure is idling high for a long period of time. The benefit to using MaxiJet's is that they DO HAVE internal thermal sensors to shut them down, rather than burn them up as some pumps we have tried. Higher heat, higher resistance, higher electrical consumption! You are correct about SOME pumps that REQUIRE Head Pressure to run more efficiently because of their design. Although not seen much in aquaria usage, worm drive and screw drive pumps often need a 'load' on them to function efficiently. Some worm drive pumps without a 'load' will self-destruct from backlashing of the gears. Magnetic driven pumps one would think would not be affected at all by head pressure, because there is no interaction between the impeller and the engine, which is just an electromagnet being pulsed to drive the core which spins the impeller. But under a heavy load, the core heats up, which in turn causes the winding driving it to heat up and under enough load, it will get hot enough to trip the thermal sensor (if your pump has one). If the load (backpressure) is too great, the magnetics breaks down and the engine cannot overcome the load on the core. Most magnetic driven pumps have floating impellers, you will often hear them chatter as you start up the pump. This is to prevent the magnetics breakdown as the pump starts and keep start up heat to a minimum. If you rigidly affix the impeller to the core, you will find that most magnetic pumps will keep losing their magnetics hold on the core as they try to start and in some cases may not start at all. In essence, an alternator works the same way, load it down and it will heat up. TTUL Gary ross@rossbagley.com (Ross Bagley) verbositized: >[...snip...] > [quoted text clipped - 31 lines] >"Security is mostly a superstition. It does not exist in nature... >Life is either a daring adventure or nothing." -- Helen Keller >Hi Ross > [quoted text clipped - 3 lines] >It only makes sense, that to do MORE work, will require MORE energy >and produce MORE heat. This is exactly what you should try. You will find that the more you restrict the flow the less current you will draw. Did this experiment in physics myself much to the surprise of the teacher. Both with air and with water pumps. Gradually reducing the flow of either dropped the amps to about 1/3 the original when stopped completely. Work is equal to force times distance. Close the valve you still have the force but not the distance. This goes contrary to the way we think of work because when you put a resistance on us we work harder but not really our internal work is greater and we sweat and heat up but the work being done on the object is actually less. If you doubt me, just restrict the flow to your pump and listen to the motor. It will speed up. How can a electric motor with out any governor be working harder but speed up. It can't. If you still doubt me, get and amp meter. I can still see the teachers face. :) Ct Midnite PS. The only exception to this would be a positive displacement pump. If you restrict the flow of these they will either stop turning or blow up your lines. They have to have an escape valve like Marcs tee but centrifugal pumps do not. http://www.geocities.com/ctmidnite53/ Hi CT Your were obtaining correct measurements but inherintly FALSE readings do to an effect called CAVITATION! Where CAVITATION is present, WORK ceases(or is reduced), the LOAD is reduced due to Cavitation, ergo current drop is eminent. It's akin to going uphill in your car, if you restrict the vehicles upward movement to the point the drive wheels lose traction and begin to spin on the pavement, the horsepower consumed will decrease due to less friction between the tires and road, but the WORK will also be creatly reduced if not ceased entirely as on ice, and you could actually begin to slide backwards down the hill for loss of traction. I can guarantee you that increasing the load on a pump increases its power consumption. If you do not show an increase in power consumption, then you have not increased the load the pump is doing. If increasing the load the engine must do, decreased the current or horsepower needed, then we would need a 450 Cummings engine to drive a wristwatch and a hearing aid battery to power an aircraft carrier. TTUL Gary I certainly don't want to get into a flame war. Please believe me when I say that your take on this is the take the vast majority of peoples have. In my class it was 24 to 2 in your favor. But it is wrong. >I can guarantee you that increasing the load on a pump increases its >power consumption. If you do not show an increase in power >consumption, then you have not increased the load the pump is doing. You are exactly right here. This is the misunderstanding. By restricting the water flow you do not increase the load on the pump. You decrease it. Restricting flow does not increase load, more water flow increases load. Work really is equal to force times distance. The more you move a given distance the more work is done. The less you move a given distance the less work is done. More water, more work. Less water, less work. Please before you really start calling me names and questioning my intelligence put the amp meter on the pump and gradually decrease the flow. You will be quite surprised. It becomes easier to see with a gas pump. I have a 3 hp cent pump to fill my 1000 gal field sprayer. With the valve fully opened and max water through it the motor is pulled down to a fraction of it's no load speed and obviously working very hard. You start shutting the valve and the motor just keeps easing up until it's just running like it's hooked up to nothing. It's pretty dramatic the difference. If you shut off the water flow out of a pump the work it's doing is close to zero. Only heating the water in side the pump a little from moving the water around and pressure but no work is being done. Your pump may burn up because it was designed to run at a lower rpm or it uses the water as a coolant but not because it's being over worked. Honest. I wouldn't kid my fellow friends on the group. :) Ct Midnite If you want a easy test of what I'm saying go get out your canister vacuum cleaner. Put you hand over the tube. The motor will speed up. And not because it's doing more work. Because it's doing less. >Hi CT > [quoted text clipped - 20 lines] >TTUL >Gary http://www.geocities.com/ctmidnite53/ For anyone else reading these posts don't think that anything I'm saying about energy used negates Marc's design with his tee and running it back into the sump. It's probably the way to go. Much less pressure in the system I would assume means much less damage to the organisms in the water. The less the pressure differences across cell membranes, the less likely they will rupture. And the tee system would give you the lowest pressure possible for any one point in the system. But it's not to reduce work on the pump. Ct Midnite http://www.geocities.com/ctmidnite53/ Unless your system is providing the necessary amount of back pressure from head alone, to make the pump run within its performance curve, then the pump is not operating efficiently and you will not get the flowrates the pump is rated to provide. In this case throttling the pump somewhat will increase its efficiency. I agree with the previous comments ... you need a proper pump curve. JCB > For anyone else reading these posts don't think that anything I'm > saying about energy used negates Marc's design with his tee and [quoted text clipped - 7 lines] > > But it's not to reduce work on the pump. Now you're talking my kind of talk. This was my "first alternate" in the list of possible reasons to tee off the return line back to the sump. Reducing the pressure inside the impeller chamber seems likely to reduce the mortality of phytoplankton passing through the pump. Which is a very good thing IMHO. This may also really simplify the experimental apparatus. I'm not really all that concerned about imparted heat. Though I will need to buy/borrow a decent microscope to count phytoplankton populations in the water... Regards, Ross -- Ross Bagley http://rossbagley.com/rba "Security is mostly a superstition. It does not exist in nature... Life is either a daring adventure or nothing." -- Helen Keller |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.