Welcome to Koi Forum. Is this your first visit? Register
Results 1 to 15 of 15
  1. #1
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746

    Solving the Gravity fed Drum filter yo-yo of water level

    Observation:

    The water level in the pond seems to oscillate between two levels, making auto popups fill and then overflow loosing a lot of water. It's also annoying visually.


    Problem/Cause:

    In a gravity fed filter, the pump draws water out of the filter - the gravity feed then fills the missing water to balance the water level. Typically the pump is at the end of the filtration just before the return to the pond at the lowest point so the pump does not run dry and destroy it - it also means the pump only pumps clean water and requires minimal maintenance.

    When a gravity-fed drum operates it allows a flow rate through that drops as the drum becomes clogged with waste it's filtering out. The water level on the clean side drops as the waste buildup reduces the flow needed for the constant pumped rate of water to be replaced. A sensor detects a low point in the clean side and triggers a wash. The wash clears the drum and the flow rate returns to the maximum again. This is the operating cycle of a drum filter 24/7.

    So if we have a complete filter with a volume of 200 litres, when the drum fills with waste and it's flow rate reduces, as the clean side drops, the pump actually drains the 200 litres of filter volume into the pond. The waste in the drum now holds back the water so the water volume is now +200 litres above normal level - this is seen as a rise in pond level. How high depends on the pond surface area so this may be 2 inches of water level rise for example.

    So from a distance the pond seems to rise and fall 2" repeatedly. However the impact is far worse:
    a) The filter system will see a rise and fall of water levels as it's water is pumped out.
    b) If the water level at rest when the pump is off and the water levels are all equalised, the water level of the pond is set just below the level of the waste chute or overflow. As the pond level rises the pond then looses water over the overflow/wastechute. In our example 200 litres goes down the drain. Then the drum cleans and the water level drops 2". If the pond has an automatic top-up set to the pond level at rest as the water level drops by 2" due to the missing water.. it refills ... only to loose the water in the next cycle - leading to higher water bills plus a topup is usually on a dechlorinator that reduces the lifetime of the cartridge.
    c) If the pump is running without equal pressure (pond pressure vs the now very empty clean side water level) then it's going to be under strain and may fail quicker.
    d) For windowed ponds the cycle can cause unsightly water marks over the 2 inches the water rises and falls.
    e) For wall skimmers with a weir system that's good - but it may cause the skimmer on a separate circuit (that may be connected to water features) to yo-yo in water flow rate too (the pressure helps the pump push more water through the water blade for example).
    f) It is annoying as hell with the water level in the pond yo-yoing..

    So the main symptom driver here is that flow rate is reduced to the point the flow rate can't keep up with the pump.

    The main cause is that the drum can't keep up with the rate of waste - causing the flow rate to drop rapidly.
    If the drum could maintain the pump flow rate with a drum 50% full of waste in a 30 minute period before the flow rate degraded then for someone that wants a 30-60 minute cycle, this would be the right drum. The rate of waste entering the drum that it filters dictates the rate of build up and the physical size of the drum dictates how much waste can be held before the drum flow rate deteriorates. Note that the rate of waste is not the flow rate - if you have a heavily stocked pond, or feed heavily, the amount of waste per unit of flow rate is higher. So heavily stocked pond will have more waste per hour in 20,000 litres filtered in that hour, vs a low stocking in that same 20,000 litres. Also having a skimmer second circuit also reduces waste rate to the drum.
    So when I say waste rate I mean the grammes per hour of waste you have coming in filling the drum.

    In reality you end up with a graph - on one axis the amount of waste in the drum, on the other axis the flow rate of water passing through the drum. Being scientific:
    1. put an ultrasonic flow meter on the output of the of the pump for a yo-yo cycle (shows the flow rate as the pump head changes with the water level) - this only needs to be done enough times for an average to find the flow rate over the yo-yo cycle. For simplicity we'll assume this now remain stable.
    2. Then measure the inlet of the drum showing the flow rate of the drum over the cycles of yo-yo for days and nights, this records the flushes over the feeding regime - the you know how much you feed in grams - assuming a waste is close to the amount of feed over that period.
    3. Vary the feeding amount.

    Then your graph should show the feeding rate vs flow rate.. and allow you to extrapolate how the drum will cope as the koi grow..

    It's not a given that incoming flow rate and waste rate are the linked - I can increase to a higher flow rate on low stocking but by waste rate still remains the same as the fish still create waste at the same rate - only the speed at which that waste presented to the drum changes.

    As a smaller micron mesh will filter more waste - it will also increase your waste rate.

    You may be able to guesstimate based on the surface area of the drum under water and the micron mesh size - if you have a waste rate of X then you will block Y of that mesh depending on the waste rate. This means you can see how quickly the mesh will clog for your pond.

    Solution(s):

    a) Reduce the depth of the trigger sensor - this increases drum washes but would reduce yo-yo in a current system.

    b) Get a drum that copes with the rate of waste YOUR pond generates and the wash cycle you want. This will allow you to set the sensor at a position that has minimal drop before triggering - thus reducing the yo-yo effect.
    One way to reduce the waste rate extracted (it means the unfiltered waste remains in the pond) is to increase the micron mesh size. Having a pre-filter with a large micron mesh helps reduce the waste rate before it gets to the drum.

    c) put a gravity fed sieve in parallel in the drum - the important piece here is the weir and that the bypass is provided to the rest of the filtration (to ensure it doesn't drain) without returning to the drum clean side which will then prevent it from flushing.
    Any overflow caused by the pump emptying the filter would then simply flow over the the sieve weir (a max is set so the weir can't continue to rise and block it).
    It does require the drum essentially have a non-return (not a valve but a high return into the remaining filter.

    Actually the drum dirty side trigger also works on this - so it can detect if the drum is being bypassed and trigger a clean - as the drum bypass is only for this purpose and not for increased flow rate - then you don't need to worry about the back flow into the clean side from the output of the sieve The sieved bypass then can be fed straight into the filter. However this will still cause the drum to wash more often - which really points to the sizing of the drum being wrong (you're adding more mesh for the given waste rate).


    Hmm got to think about this more


    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  2. Thanks Jampot, jayjt29, blondie Thanked / Liked this Post
  3. #2
    Senior Member Rank = Jussai g mac's Avatar
    Join Date
    May 2020
    Posts
    1,431
    Thanks / Likes
    2641
    This has made my head hurt, caffeine is needed and then I'll read it again

  4. Thanks freddyboy, Scamp, blondie Thanked / Liked this Post
  5. #3
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746
    Quote Originally Posted by g mac View Post
    This has made my head hurt, caffeine is needed and then I'll read it again
    Yep.. the TL;DR answer is: if you see yo-yo then your drum is too small for the amount of poop the fish are making (regardless of flow rate).
    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  6. Thanks Jampot, blondie Thanked / Liked this Post
  7. #4
    Senior Member Rank = Gosai Jampot's Avatar
    Join Date
    May 2020
    Location
    Chorley. Lancs
    Posts
    251
    Thanks / Likes
    308
    if you see yo-yo then your drum is too small for the amount of poop the fish are making
    I like this version better

    Nice write up, in your inimitable fashion Nick.

    Thanks

    Jim
    I don't keep fish, I keep water. I don't keep fish, I keep water. I don't keep fish I kee........

  8. #5
    Basically, as your screen blocks the flow reduces and the pump returns more water than the drum can provide.
    “Get a bigger drum” is a bit of a piss off though, I’m guessing most, like me, on 4500 gallons a “35” sized drum would be ample however, I’ve said before (now I know about these things) manufacturers are extremely liberal with their flow rates and yes, when the screen is clean you may be able to flow what they quote however, when it’s dirty you certainly won’t.
    I would say based on my own experience it’s probably 60% of the maximum give or take.
    I defiantly don’t have the confidence to run my 2x20k’s much over 60% and if I went away for any length of time I’d be turning them down a couple of notches.


    Sent from my iPhone using Tapatalk

  9. Thanks Scamp, blondie Thanked / Liked this Post
  10. #6
    Good explanation by the way Nick Solving the Gravity fed Drum filter yo-yo of water level


    Sent from my iPhone using Tapatalk

  11. #7
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746
    Quote Originally Posted by bigcarpchaser View Post
    Basically, as your screen blocks the flow reduces and the pump returns more water than the drum can provide.
    “Get a bigger drum” is a bit of a piss off though, I’m guessing most, like me, on 4500 gallons a “35” sized drum would be ample however, I’ve said before (now I know about these things) manufacturers are extremely liberal with their flow rates and yes, when the screen is clean you may be able to flow what they quote however, when it’s dirty you certainly won’t.
    I would say based on my own experience it’s probably 60% of the maximum give or take.
    I defiantly don’t have the confidence to run my 2x20k’s much over 60% and if I went away for any length of time I’d be turning them down a couple of notches.


    Sent from my iPhone using Tapatalk
    The get a bigger drum - what you're doing by adding a prefilter or a bypass filter is that that you're adding mesh surface area agains the incoming waste. So if a bigger drum has a larger mesh that solves the problem - until you see a couple more Koi that you want...

    I agree. Manufacturers should state the wet drum open surface area - this area changes with micron size, the weave of the mesh and the mesh support. It defines the waste handling and maximum flow. Although mesh manufacturers state the open area as a percentage of the mesh, so drum manufacturers - this really should be in cm2 or m2.

    If you have a 20,000 lph clean then the mesh surface area will support that, however scale, bio film etc all reduces that open area and reduce the max flow rate. I tend to agree that a rough rule of thumb being 60%. However this also depends how often is acceptable for a drum to wash? A drum sized for 20000lph washing continuously vs 20000lph washing every 30-60 minutes all depends on the amount of waste building up and the rate of waste that drives the wash rate. Or simply - you want something like want 20Klph rate, washing every hour with X grammes of waste per hour = required drum open mesh area.

    Next question is - flow rate.. why is it important to the pond anyway? Surely waste rate is more important?
    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  12. Thanks blondie Thanked / Liked this Post
  13. #8
    Surely they’re either one and the same or interlinked.
    You can’t get the waste to the drum without flow, it won’t get there by itself lol


    Sent from my iPhone using Tapatalk

  14. #9
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746
    Quote Originally Posted by bigcarpchaser View Post
    Surely they’re either one and the same or interlinked.
    You can’t get the waste to the drum without flow, it won’t get there by itself lol


    Sent from my iPhone using Tapatalk
    Almost - if your koi eat 500g of scoobie snax.. the resulting mess into the same 10,000lph is going to be higher that if you had less koi eating just 100g.

    Or another way.. if you have a stream. A heard of elephants takes a dump into it then the next day a couple of humming birds poop into it.. Same flow rate. but if you're the stream fisherman with a net stretched across the experience is going to be very different!
    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  15. #10
    But your slow flow rate might leave some elephant poo behind?


    Sent from my iPhone using Tapatalk

  16. #11
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746
    Quote Originally Posted by bigcarpchaser View Post
    But your slow flow rate might leave some elephant poo behind?
    Well in a koi pond you want to keep the waste mobile but also remove it asap from the pond?
    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  17. #12
    Ergo...my drum will keep overflowing until I get a bigger one or I have a change around somehow.


    Sent from my iPhone using Tapatalk

  18. #13
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746
    Quote Originally Posted by bigcarpchaser View Post
    Ergo...my drum will keep overflowing until I get a bigger one or I have a change around somehow.
    Yup - all roads lead to Rome.
    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  19. Thanks freddyboy Thanked / Liked this Post
  20. #14
    Or in this case (koi pond), bankruptcy lol


    Sent from my iPhone using Tapatalk

  21. Thanks freddyboy, NickK-UK, blondie Thanked / Liked this Post
  22. #15
    Extreme Koi Member Rank = Adult Champion NickK-UK's Avatar
    Join Date
    May 2015
    Posts
    1,539
    Thanks / Likes
    1746
    An idea has just occurred to me but it requires replacing the variable pump controller.

    Currently we have constant flow rate variable pond water level. One option is constant pond level variable rate.

    If the drum water level is an issue - then varying the pump speed would ensure that the system doesn't have a yo-yo. Naturally it is simply reducing the flow and therefore the filtration to cope with the drum limitation. However it could offer a stable water level. Naturally this means that the drum would never sense the drop on the clean side.

    The system monitors the water level, on the inlet side, if the water starts rising the controller can reduce the pump speed. Then the controller can look at the rate of speed and trigger the drum before the forecast flow rate drops below the configured rate.

    This flow rate drop doesn't have to be massive - just enough to get a clean every 20 minutes for example without staving the filter bacteria or impacting the pond water stats.

    You'd need a variable frequency inverter to replace the 24V pump controller, a ardiuno to oversee and control things and a couple of water level sensors (one a "max" fail safe and the other senses position). If the max sensor is triggered then the system automatically triggers a wash. In the end it's probably cheaper to buy an upgraded drum with a larger mesh capacity.


    A simple version with constant rate pumping is run two sensors - a rise (using a pump-fed configuration in the gravity fed system) wash sensor in the inlet and a drop (gravity-fed) wash sensor that both trigger a wash. That way you can limit the yo-yo (assuming the drum is the only component making it yo-yo) and the system will also trigger normally with the water level drop - this takes into account water volume changes (evaporation etc).


    I'm, still in the mind that a drum is superb in what it does but the pond/filtration as a whole needs careful design - simply having variable valve (ie the drum) at one end of your filter and a pump at the other isn't perhaps an easy design to get right. PLUS if you have have additional filter loops - this add additional variables. Not impossible but not something you can statically manage by setting a valve or pump and the leaving it for years. Interesting a pump fed configuration already has this variable flow - the pump feeds constant, the water levels in the bio and filter change but the water rate out will cut off at the point the water level drops below the return, so the only yo-yo used is in the drum inlet chamber that triggers the wash. If the pump stops and the system back flows through the pump then the pond level rises and possibly overflows. However that's in a failure scenario.
    14000l, my mutts: Chargoi (2010), Doitsu (2022), Tancho (2022), Kujaku (2022), Hi Utusri (2022)

  23. Thanks blondie, RG1X Thanked / Liked this Post
 

 

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
All times are GMT +1. The time now is 02:17 PM. Online Koi Mag Forum
Powered by vBulletin® Version 4.2.3
Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.

vBulletin Improved By vBFoster® (Lite Version), © UltimateScheme, Ltd.