Gate or Ball valves?

[Piscevore]

I'm going to have my overflow drain pipe Y out and feed two locations at once under the new tank. I would like to put a valve before this Y if I want to slow the total amount of gravity fed water, or stop it. I am also planning on a few other locations in my tank where I would like to be able to simply shut off water flow, not necessarily limit.

Ball valves are the only kind available at my local HD, and I so far what I have read about gate valves suggests they are best used in the on or off position, nothing in between (plumbing supplier site).
I've seen many people writing about the NEED to use gate valves with no explanation... I know I'll have to pay more for the gates, and probably order them instead of local pickup, but I'm not taking that into consideration in the comparison. I just want to know side by side which is more desireable, or if they are better suited to different applications. I am concerned most about reliability (no failure!)

 

[ScottM]

True gate valves allow you to control the flow better than a ball valve.
ball valves have always worked best for me. i think loews is cheaper on ball valves.
if you want to i am placing an order with us plastics soon. check them out:
www.usplastics.com

 

[mikrok]

I wouldn't use a gate on a high pressure application...I think they're better of for precision. In your case, I think either could work fine, since you won't have back-pressure

 

[Matt L.]

There is a common misconception in the hobby that gate valves are better at throttling flow than ball valves. This is not really true, and the plumbing supply site did give you the (more) correct statement: overall, ball valves are much better for throttling flow than gate valves. The CV versus valve position curve for a ball valve is much more linear than that for a gate valve.

There was a thread on RC that I participated in back in February 2004, where I discussed the hydraulics behind my assertion. After all, I wouldn't want any of you going up to a hydraulic engineer, asserting that gate valves are better for throttling flow than ball valves, and getting laughed at.

I don't know where the misconception originated, or how it could basically be 100% backwards, but its out there.

Yet there is some truth behind myth in this hobby that gate valves are better at throttling flow than ball valves. For the valves available to us in this hobby, and the small diameters we use, it appears (to me) that gate valves might be better at throttling flow than ball valves, but all this really means is that the Home Depot ball valves are really, really poor*.

Personally, I would not use a gate valve to throttle flow -- I would just look for a higher quality ball valve (or globe valve); perhaps one with some AWWA merit to it. Of course, these will cost far more than the $5 valve you buy at Home Depot.

I'm going to have my overflow drain pipe Y out and feed two locations at once under the new tank. I would like to put a valve before this Y if I want to slow the total amount of gravity fed water, or stop it. I am also planning on a few other locations in my tank where I would like to be able to simply shut off water flow, not necessarily limit.

I am very much opposed to throttling gravity fed drain water, although many people do it succesfully. I think it is because that configuration is poor hydraulic design, and any objection I have carries over from my design engineer days. Are these two separate locations you wish to drain to hydraulically connected?

On the other hand, putting on/off valves in your lines is very good hydraulic engineering. All my lines have a ball valve for on/off, thereby isolating sections of my plumbing. It has been invaluable for pump maintenance, etc. You may see more when you stop by today,

Matt

* I'm thinking of another four-letter word, but this is a family board.

 

[Piscevore]

Matt &co,

I was thinking of reducing flow on the gravity feed if for example the main return pump was damaged or replaced due to failure and unable to return as much as was draining. At the least I'd want be able to stop the flow, at best limit.

I'm trying to plan as far ahead as possible ;-) I haven't finalized my sump plans, and that has a lot to do with what I'll be using for plumbing materials. I'm going to bookmark US plastics though, because they've clearly got everything, and I'm willing to bet its higher quality than what I'll find at my local shops.

 

[joefitz]

I'm not sure I'm following this...you are talking about throttling the water going from the main tank into a sump? This doesn't make any sense to me -- I can't see any reason why this would ever be a good thing. If the return pump fails then water will no longer drain from the main tank to the sump (at least not after enough water drains to put the water line below the overflow teeth).

A properly designed system (ie, flood "proof") would never require or desire a restriction on the flow of water from the main tank to the sump. It can only drain as fast as the water is being returned from the sump to the main tank. If you slow the flow down from the main tank then the water level in the sump will continually decrease while the main tank will continually increase until either the sump pump fails (runs dry) or the main tank overflows/floods.

I hope I'm just missing the entire point of this conversation which is entirely possible, in which case I apologize in advance.

 

[Piscevore]

No no, you're right. I think I'm being over cautious.

I'm afraid that when water -is- coming out, TOO much might come out at some point. That will be properly avoided if I plan it right. This is basically boneheaded-planning contigency plans

For instance if: I crack my -glass- sump, I don't want to wait for 20 gallons to drain before it stops

Or if like I said, my pumping ability suddenly decreases, but I want to be able to continue making exchange in the tank, it would be necessary to control the OUT flow from the tank. Seems to me its equally as dumb to not have a flow control where you need it and just rely on a "fool proof" plan, as it is to "just" put a valve before and after everything. I'm looking for some balance, and I appreciate the input!

 

[Matt L.]

I'm not sure I'm following this...you are talking about throttling the water going from the main tank into a sump? This doesn't make any sense to me -- I can't see any reason why this would ever be a good thing. If the return pump fails then water will no longer drain from the main tank to the sump (at least not after enough water drains to put the water line below the overflow teeth).

A properly designed system (ie, flood "proof") would never require or desire a restriction on the flow of water from the main tank to the sump. It can only drain as fast as the water is being returned from the sump to the main tank. If you slow the flow down from the main tank then the water level in the sump will continually decrease while the main tank will continually increase until either the sump pump fails (runs dry) or the main tank overflows/floods.

I hope I'm just missing the entire point of this conversation which is entirely possible, in which case I apologize in advance.

Joe,

That's a very nice amd concise way to phrase it. I'm impressed.

In short, I feel an isolation valve on the return line is a good idea, but a throttling valve on the retrun line is a bad to unnecessary idea,

Matt

 

[joefitz]

I hear what you're saying Piscevore with respect to contingency planning; however, what I'm suggesting is that throttling the flow from the overflow -- that is, from the main tank to the sump -- is NEVER a good idea. Believe me, I was in your exact same position trying to figure out everything I could do to minimize all flood risks...living on the 8th floor of an apartment building I was very concerned about flood damage. But what you are suggesting will only increase the risk of flooding, throttling the flow (as opposed to shutting it off completely) can NEVER decrease the flood risk.

Being able to completely stop the flow could theoretically be useful in the sense that you could avoid draining some amount of water from the main tank into the sump (presumably if there were some sort of leak in the sump this would be useful). But there is never a situation where throttling the flow is necessary or desireable -- if the main pump starts pumping at a slower rate for some reason then the flow from the main tank through the overflow and into the sump will slow in the exact same manner at the exact same time. The pump controls the flow rate from the tank, through the overflow and back to the sump (unless the drain line is undersized and cannot flow as fast as the pump in which case you have an entirely different problem with a different solution).

Personally, I probably wouldn't put anything on the drain lines. But if I did, I would do exactly as Matt suggested and put some sort of an isolation/ball valve at the base of the return line to the sump, and I would use it either all open or all closed, never anywhere in between.

By the way, the reason I'm suggesting putting the ball valve as close to the sump as possible (and I would suggest potentially having it be within the sump enclosure provided you have enough room and easy access) is because this will allow you to maximize the water being held outside of the sump (in the main tank and the plumbing) and therefore minimize the actual sump volume. Since the only use for this valve occurs when the sump fails, the assumption is that you'll have to empty the sump (in all likelihood completely) so you might as well have as much water in the main tank and plumbing as possible (that much less water to pump/move into a separate container).

Lastly, I highly recommend determining exactly how much water will drain into your sump when the pump fails. If your system is already set up you can just shut the pump off and see what happens. Make sure you shut off any other devices that also contain water that would otherwise drain into the sump, such as a protien skimmer or potentially a calcium reactor... If the system isn't set up yet, you could get a very accurate estimate of the water volume with some basic calculations and some legwork to estimate how much water your drain pipes hold when they are full of water.

Once you know how much water will drain into your sump you can a) ensure your sump is large enough to handle the water volume and b) think about other ways you might be able to add some redundancy to protect against flooding (perhaps putting the entire sump within a large plastic container or even sealing the entire base of the stand so that it could hold water).

I'm not sure how clear I've been but I hope you get what I'm trying to say. Personally, I don't think it is worthwhile to be able to close the drain line because if there is a catastrophic failure of the sump then you'd have to be there (and watching the tank) to be able to shut the valve and even then, you are probably not talking about much additional water volume. If there is a slow sump leak (much more likely), you'll have to drain the sump one way or another so shutting the valve completely and turning the pump off (you have to do both, one after the other) won't really save much time.

Anyway, please don't take anything I've said as ridicule or anything other than friendly advice from someone who was in the exact same boat you are in as little as a year ago. Good luck and please don't hesitate to ask questions if you have any.

Joe

 

[mikrok]

I guess I didn't understand what your purpose of the Y feeding two locations would serve. then i have a question for you, matt & joe (I hope i'm not hijacking this thread, but expanding on it) - I plan on expanding my sump/fuge. I have a 58 with a 15 gallon sump and hang on fuge under it. What i'd like to do is build an extension off of my stand to the left and put a 20long as a sump. this would house my ASM skimmer. Under the stand would be a 20high refugium. This would be higher than the 20long sump which would be on the floor, so my plan was to split my tank drain line. (The fuge would be gravity fed, and would be drilled to allow a gravity return to the sump which would be slightly below it, right next to it.) A Y fitting with a ball or gate valve only on ONE side - the side to the fuge to only get a small amount of flow. The other side of the split from the drain would be 100% to my sump. Do you guys see any problems with that? If something were to happen, I imagine all flow would go to my sump which would be able to handle all flow anyways.

I figured you wouldn't actually be cutting down flow to your sump, piscevore, or I would have also said don't risk slightly closing a valve on your drain. things can get stuck in there like snails and what-not.

 

[joefitz]

Mikrok, I think what you're suggesting is fine (if I'm understanding it correctly). Basically you just want to divert some flow from the drain line into your fuge with the rest of the water going straight to the sump. You then have the fuge set to drain to the sump. Sounds like a plan to me. The only word of caution I would give is that I would make the fuge able to handle the entire flow in case somehow the drain to the sump gets clogged -- ensuring the fuge's drain to the sump can flow at the full drain rate will eliminate any risk of flooding.

As an FYI that seems a touch related I'll mention another idea I found from a fellow reefer at some point in the past. Plumb your return line with a tee/wye. make the stright-through portion of the tee go from the pump on one side to the rest of the return plumbing on the other side. The short side should be sent to a ball valve and then directed back to your sump. This ensure your pump can always run at 100% capacity which is much healthier on its motor. In addition, it can allow you to fine tune/adjust the flow going into the tank if you have some reason to do so.

 

[mikrok]

excellent - thanks joe! I understand about the tee and the valve - great for dialing flow without the backpressure.

 

[Piscevore]

Joe- Beautiful. I failed to think the thought completely through. You're absolutely correct (again). I was thinking from a full tank perspective, but once it drains it will equalize with the return pump...Elementary!

As far as the tank, its not even drilled yet for the overflow...I have calculated the needed overflow length and drain diameter on reef central. Now I have to adjust the height of my overflow to determine what the maximum drainage would be in the case of the return shutting off. I have put off doing the math because I've got a bow tank, not a square tank It requires a little bit of geometery (or calculus). I'm pretty sure I'll be using a 29 gallon sump, so that gives me a max volume and thus dictates the maximum "bite" I can take.

As far as what mikrok is saying, that is exactly what I've got planned, a "y" taking some flow to a fuge, and the rest to the skimmer. This very much helped me figure out what I need to do. Thanks for the great information guys!

 

[Matt L.]

Not to pick nits, but...

...This ensure your pump can always run at 100% capacity which is much healthier on its motor...

Believe it or not, this statement isn't exactly true either.

In explaining this, I will make the assumption that the pump we are talking about is an ideal centrifugal machine. Often, less expensive powerheads and the such kind of fall outside ideal centrifugal machinery, and it could be argued that the affinity laws* do not apply to them.

However, for centrifugal machinery, the most efficient operating point on the pump curve is somewhere in the middle between the zero flow/maximum head (shut-off) and the maximum flow/minimum head (run-out). Recall that efficiency is the ratio of power actually gained by the fluid to the shaft power supplied.

It is actually not a good idea to run your pump at or near the run-out condition as this errant, fellow reefer does: it is not better for the motor, and you're wasting energy. Moreover, if there should be a sudden reduction in hydraulic head, the pump will travel past the maximum flow/minimum head condition, which is actually very bad for the motor.

The following is an idealized pump performance curve, showing where a centrifugal pump should be operated: at its best efficiency point (BEP) shown in blue. The concentric circles of 50%, 55%, and 60% are efficiencies:

The Engineering Toolbox, which is where this image is taken from, is a great resource. I highly recommend that everyone who has questions about hydraulics and fluid mechanics take the time to read through there.

If the above image is difficult to understand, or you would like a greater understanding, you may study this curve as well.

I hope this helps,

Matt

* The Affinity Laws may be seen, written out in equation form, here.

 

[joefitz]

Very interesting. I'm not sure I completely follow the graph(s) though.

Is it suggesting that our pumps should be run at a rate less than "full"? Or that our pumps are more efficient when there is head pressure? Or neither or both?

From what I can make out, it appears the curve that the blue dot is on represents the "ideal" (most efficient) operation. The 10/8/6 in dia. curves (concentric circles) represent the diameter of the pipe? And the 50/55/60 curvs (concentric circles) represent the efficiencies (with the rightmost intersection with the "ideal" curve representing the highest efficiency). Is that right or am I off base? And how can we apply that graph to our own pumps?

It appears to be suggesting that some amount of head pressure is necessary to make the pumps operate most efficiently... I would think the amount of head required would greatly depend on the pump itself and its operating characteristics. Some pump seem to be designed for high head situations while others are not so I would think the amount of head pressure required to operate efficiently would be quite different for these pumps. Could it generally be stated that most pumps we use in our systems probably have enough head introduced by our return feeds and plumbing that there is no need to "artificially" increase the head pressure (by potentially throttling the pump)?

And then there is the "practical" matter of the application of the "ideal" to the real world. Do you think it is a bad practice for us to be running our pumps at full throttle?

 

[Matt L.]

What are nits anyways

...Is it suggesting that our pumps should be run at a rate less than "full"?

Yes, although the curve isn't making that suggestion. The curve simply tells you the optimal point at which to run a pump. Engineering suggests that you run a pump around its best efficiency point.

Bear in mind that each pump has its own performance curve, and performance curves are like finger prints or snowflakes; no two are alike. Sometimes, the best efficiency point is closer to maximum flow. Other times it is closer to maximum head. However, because the pump wasn't designed by monkeys, the designers probably put the best efficiency point (BEP) at a practical location on the curve.

...Or that our pumps are more efficient when there is head pressure?

Yes, although again, that one curve shown is just representative. Due to the Affinity laws, and the design intent behind the pump, pumps are always most efficient when some head is applied.

...From what I can make out, it appears the curve that the blue dot is on represents the "ideal" (most efficient) operation...

Yes, although it is a little more complicated than that. The curve that is concave up represents the system curve; the curve corresponding to head loss through the plumbing versus flow rate. This curve will intersect the pump curve, and as you can see, the pump was chosen correctly in that its unique pump curve intersects the system curve near the best efficiency. In practice, we are rarely that lucky.

...The 10/8/6 in dia. curves (concentric circles) represent the diameter of the pipe?

No, you probably should ignore these. These concentric curves show how the pump curve changes when you change the diameter of the impeller. Each of those curves is like for an impeller that is an identical, yet scaled down version of the other. The 10", 8", 6" are impeller diameters.

However, you will notice a dotted curve that is concave upwards. This curve shows what would happen if you changed the plumbing in your system so that the head loss is greater. Say you have a throttling valve (ball valve, of course) on the pump discharge, if you throttle the valve more closed, the system curve changes, and it intersects the performance curve of the pump at a different point.

And the 50/55/60 curvs (concentric circles) represent the efficiencies (with the rightmost intersection with the "ideal" curve representing the highest efficiency). Is that right or am I off base?

Yes, you are right. There is a 45%, 40%, etc... curve outside the 50% one, but these aren't shown.

And how can we apply that graph to our own pumps?

Well, again it is hard to use that specific graph for anything other than a general understanding of how centrifugal pumps operate, and the many variables involved in their operation. Suffice it to say, if the manufacturer has a performance curve available, you will estimate the head of your system curve, and make sure that the system curve head is not too close to the maximum discharge or maximum flow condition. Without best efficiency curves, it is hard to make any further decision. No pump manufacturer that I have come across gives out efficiency curves.

It appears to be suggesting that some amount of head pressure is necessary to make the pumps operate most efficiently...

Yes. Again, pump efficiency is the ratio of power actually gained by the fluid to the shaft power supplied.

I would think the amount of head required would greatly depend on the pump itself and its operating characteristics. Some pump seem to be designed for high head situations while others are not so I would think the amount of head pressure required to operate efficiently would be quite different for these pumps. Could it generally be stated that most pumps we use in our systems probably have enough head introduced by our return feeds and plumbing that there is no need to "artificially" increase the head pressure (by potentially throttling the pump)?

Yes, again, remember that each pump is unique. The manufacturers have designed the impeller for a specific set of operating conditions. Also, must of us have enough head in our systems that unless you were using a high head/low flow pump for a system with very little discharge head, you will be okay, and even then, these motors are resilient.

And then there is the "practical" matter of the application of the "ideal" to the real world. Do you think it is a bad practice for us to be running our pumps at full throttle?

Again, it depends on the pump, but I would say probably not. The manufacturers have designed these pumps for reef use. Still, look at the pump curve, the closer you are to the middle of that curve, the better off you are.

Matt

 

[joefitz]

Hey Matt, thanks for the info!

Joe

 

[JohnK]

I think this is the most confusing thread I've ever read.

That said, as ususal, thanks matt for the incredibly informed responses.

as for the question about the common misconception about gate valves vs ball valves I have two answers about where that came from.

1. Inland reef's website had a write up about the difference between applications of the different valves that at least reinforced the "misconception", and I suspect that a lot of local reefers (myself included) probablly got their info there.

2. IME regardless of the engineering issues involved, gate valves allow a much more precise adjustment of flow but tend to leak in the off position. In contrast ball valves (again IME) are very hard to adjust in small incriments but do work very well for on/off applications. (ever tried to carefully adjust a skimmer's foam level with a ball valve?)

*** I am sure that everything matt said is correct at least from an engineering perspective, he clearly knows what he is talking about****

beyond all that I would agree that It is never a good idea to restrict the total flow from the overflow to the sump. If nothing else, given time and accumulation of misc debris, potential snail shells, algae exct if the total overflow volume is restricted you risk an overflow/flood from the main tank. If everything is adjusted perfectly but you get lazy for a week or a month it is quite likely that the perfectly adjusted overflow rate will end up getting reduced by some sort of fouling and you will end up with an overflowing main tank and a return pump running dry.

Not intended to argue with anyone, just adding $.02

as far as the other stuff about ideal pump conditions I have nothing to say except that Matt knows his stuff

(hope I'm not just missing the point here, I'm tired)
jk

 

[Matt L.]

...as for the question about the common misconception about gate valves vs ball valves I have two answers about where that came from.

1. Inland reef's website had a write up about the difference between applications of the different valves that at least reinforced the "misconception", and I suspect that a lot of local reefers (myself included) probablly got their info there.

2. IME regardless of the engineering issues involved, gate valves allow a much more precise adjustment of flow but tend to leak in the off position. In contrast ball valves (again IME) are very hard to adjust in small incriments but do work very well for on/off applications. (ever tried to carefully adjust a skimmer's foam level with a ball valve?)...

Thanks for the response. I will contact Inland Reef about the information on their web site, and respectfully request that they correct their suggestion, as it is unintentionally leading to much confusion.

Again, if reefers were not having the positive experience with gate valves over ball valves (for modulating flow), the fallacy that gate valves are better for adjusting flow than ball valves would not perpetuate. But I think this is simply because reefers are using poor quality ball valves.

Matt

 

[Matt L.]

I forget that there is no more Inland Reef, so I guess this information is stuck out there. Oh well,

Matt