So you (Still) Want a Crystal Clear Pond: Revisiting Undergravel Filtration

I wrote an article for the May/June 2014 issue of POND Trade magazine under a similar title: “So You Want a Crystal Clear Pond.” If you don’t recall it, check it out in the archives at before you dive into this update.

undergravel filter pond

This picture shows the completed piping for an undergravel filter. A layer of gravel should extend up to 2 inches above the top of these pipes, which run out of the pond to the suction side of a pump. Fitz’s Fish Ponds LLC built this pond. (Photo courtesy of Mike Hall)

You might notice that my 2014 article generated a few dozen comments and questions from my fellow readers, so I’m happy to take this opportunity to expand on what I wrote four years ago about undergravel filtration. I’ll also address some questions that have since come up.

First, just a quick update — my first undergravel filter is still running great and has not needed to be cleaned in the last 27 years. I’m guessing that the liner will fail before the filter needs to be cleaned or stops working.

Koi Safety Issues

I came across a Facebook post about three weeks ago with a discussion among several well-known koi experts. They were familiar with undergravel aquarium filters and felt that undergravel pond filters would behave in the same way. In other words, the filter could fail or build up toxic areas.

Could this happen in a pond undergravel filter? It’s possible, but not from the waste from the aquatic life present in the pond. It would require a lot of outside debris infiltrating the pond and simply left there to cause a toxic, rotting situation. Of course, this is true with any pond using any type of filter system. An advantage to having an undergravel filter is that debris is generally not an issue, unless it’s so large that it actually prevents the pump from being able to pull water through it. We’re talking about a layer of debris 4 to 6 inches thick. It would certainly be producing poisonous gases before it would get this thick, and you have to assume that the pump would suck it in and bring it up to the surface, where it would rise into the atmosphere before it would harm the aquatic life.

A likely response to this answer is, “That’s not why undergravel aquarium filters fail!” And that is correct. They fail because the gravel or slots in the undergravel bottom plate get clogged up. When the filter is designed as I specified in my last article, it won’t clog up. Water will continue to flow through all areas of the filter, and the filter will continue to work as advertised.

Size Matters

The next logical question is, “Why doesn’t it build up enough debris to clog it up?” Simply put, in a properly constructed system, there should be sufficient space between the gravel so that debris continues to move and break down, eventually exiting though the filter.

  But what about debris that is too large to get into the gravel? This would normally include outside environmental debris, such as leaves or sticks. This material tends to lay on the gravel surface until it is manually removed, or until it breaks down into smaller pieces. It will break down more quickly with water moving around it, rather than just sitting at the bottom of the pond. But it still could take a very long period of time. It is a good idea to have another type of filtration system, such as a skimmer, that is designed to remove floating debris from the pond. I would recommend a secondary filtration system for any outdoor pond, regardless of its primary filter system.

But how closely do you have to follow the directions for building an undergravel filter, which I outlined in the previous article? I can’t answer this one completely, as I have not branched out from my formula very much.

The farthest I have strayed from my proven design is a pond that I built 11 years ago. The homeowner wanted a pond with a completely calm surface. It was for plants primarily, but they wanted some fish, too. To meet these requirements, I designed the system with a submersible pump in a sealed box with an open pipe leading into it from half of the undergravel filter. The output from this pump went to the other half of the undergravel filter as a pressurized filter. The last thing this homeowner wanted was small gravel, and I had my doubts, but we used pea gravel.

An undergravel grid is ready for gravel.

Eight years later, I talked to the homeowner, who had done zero maintenance on the system. Their only complaint was that they occasionally had to dye the water, because the pond was too clear to show off the plants as much as they wanted. I am not crazy about this system for the average homeowner, as it uses a submersible pump, which doesn’t last as long as an external pump. Moreover, this pond doesn’t have a skimmer, so the homeowner has to remove surface debris manually. Most homeowners would not keep up with debris like this. You can see a picture of this pond in my previous article.

Add an Airlift

It may seem hard to believe, but you can filter a large pond using 60 watts of electricity, or even less. This is another huge advantage of undergravel filtration systems. Because the filter is in the pond, it can run without having to lift water above the surface of the pond. This creates the perfect scenario to implement an airlift system. With 60 watts of power operating a good air pump, you can easily move 10,000 gallons of water through the filter.

What other benefits can be expected from this kind of system? You could not have a better system for oxygenating the pond. If designed correctly, the filter should draw water from the entire bottom of the pond and bring it to the surface, where it can absorb oxygen from the air. There should be no dead areas in the bottom of the pond. Other than skimmers and gravity-fed settling chambers, there are very few other filters that can be run by airlift systems.

underground airlift system

An airlift system was integrated by Magnolia Ponds (photo courtesy of Max Taylor).

There are always going to be some who believe that undergravel systems promote the production of hydrogen sulfide. The truth is, this can happen — if the system is designed incorrectly, that is. I have found that a correctly designed system is about as maintenance-free as can be. My own pond requires fewer than 10 minutes of maintenance per month during the spring and summer. It has been running for more than six months now with no maintenance at all, and it will not need maintenance until the hoop house is removed.

Across the United States, there are hundreds of these ponds that were designed correctly and continue to work perfectly. However, there is no such thing as the perfect system. If there was one perfect system, everyone would use it, and we wouldn’t be having this discussion. But all things considered, you can be assured that if implemented per the design instructions, undergravel filtration works very well.

20 Responses to So you (Still) Want a Crystal Clear Pond: Revisiting Undergravel Filtration

  1. Jacob Spering July 29, 2018 at 12:15 PM #

    Hello! I like the idea of an airlift system in a undergravel filter pond. But how does it work? And how many gallons of air per hour would you need to move 10,000 gallons? I love your work!

  2. Mike White July 30, 2018 at 5:38 AM #

    Hi Jacob, The way an air lift works is that the rising air will push water to the surface. To move 10,000 gallons of water per hour I would recommend a 60 lpm air pump. It can be done with a 40 liter per minute air pump but everything has to be just right.

  3. Wolfgang Ebert August 1, 2018 at 10:57 AM #

    Hello Mike,
    I’m writing from Frankfurt, Germany.
    I like the system with undergravel filter very much, but do not understand it in detail.
    Are the tubes the filter or do you pump the water into a mechanical filter and then into a biological filter?
    I’m missing the filter !!!
    The most recommended way here would be to have 2-3 110 mm tubes sucking the water from the ground, bing it to the mechanical and then into a biological filter. From there pumped back into the pond.

    • Mike White September 6, 2018 at 2:51 PM #

      Hi Wolfgang,
      In an undergravel filter system the pipes used are not the filter but the collection devise used to collect the filtered water. The gravel that the water has to past thru to get to the collection pipes is the filter. Does the pipes do any filtering? The answer yes but are limited in surface area.

  4. Terri September 5, 2018 at 5:26 PM #


    Thank you for a follow up article. Your first article inspired me to enlarge my koi pond and include an under gravel filter. Do you aerate your pond or does the suction grid remove the gases and circulate the water?

  5. Mike White September 6, 2018 at 3:17 PM #

    Hi Teri,

    To answer your question yes I do aerate the pond. That is one of the ways I use to help keep the dissolved oxygen level at saturation for the health and well being of the pond and the inhabitants of the pond. With that said the suction grid can circulate water with dissolved gases to the surface.


  6. Sarah September 12, 2018 at 9:20 PM #

    Hi Mike. I have a 20 x 12 Pond about 30 inches deep. I am thinking of using 2 inch PVC pipes spaced 2 feet apart with the three holes every one foot. I just purchased an in-line submersible/above ground 8000 gallons per hour torpedo pump by Cal pumps. This will feed water to my upper level Bog pond which is 5 x 20 x 12” deep 3/8 Pea gravel made as the Nelson water gardens family suggests. in the big pond, should I use a bigger rock Like 1 inch to 3 inch to maybe even 5 inch and only use 2 inches of rock above the pipe to create better water flow? This is because I’m not filtering water down there, I’m using your system basically to collect water and pump it up to the bog and let the bog filter it. Or should I do it as you say with the 3/4 to 1 inch River rock? I’m afraid I may create too much of a filter on the bottom and starve the bog plants of nutrients. I will have a light fish load of 50 to 75 to 100 fantail goldfish.

  7. Mike White September 13, 2018 at 3:53 AM #

    Hi Sarah,

    First the plants are going to want nitrates and no matter how much filtration you have they are not going to remove nitrates. I would use the 3/4 to 1 inch gravel. The other thing I would do is I would put the 2″ pipe one foot apart and not 2 feet apart. This will keep the water cleaner which will help in not having to clean the bog filter as often.


  8. Anthony Milana September 23, 2018 at 5:52 PM #

    Hello Mike, thank you for writing both these articles!
    I am in the process of designing and building an 18’ x 12’ “natural wading pond”. This pond will consist of a 1’-2’ plant zone and a 3’ wading zone. I would like the bottom of the wading area to house an under gravel filter system. The base there will be 10’ L x 6’ W x 3’ D. With three, 3/8” holes drilled on the bottom, and at 60 degrees on either side of the bottom every 6 inches, I get an area of the holes to be 50 in^2. Would you recommend a 2.5” main line with 2” arms? Could you advise on a pump size for just this portion of the system? There will be another pump in my skimmer that will feed jets to help push the surface debris to the skimmer. This pond will be in the desert and will see summer temps around 110 F occasionally. Do you think the PVC with 2” of gravel on it could handle people walking on it?

    Thank you,

    • Mike White September 24, 2018 at 5:56 AM #

      Hi Anthony,
      As far as the size of the pump to be used I can not give you the size that I would use. I would have to know the volume of the pond and set the size of the pump based on that if it is not going to be used to run a waterfall. The 2.5″ pipe can handle a flow up to about 7000 gph. People walking on the pipe won’t hurt it at all. But it won’t be easy for people to walk on gravel on the bottom.

  9. Dan September 23, 2018 at 10:49 PM #

    Thanks for these articles!
    I’m just starting to design a double pond, one pond for koi and one for plants and a few gold fish. I plan to have them side by side with a total combined volume of 3000 gallons. I want to use airlift pumps. It appears from your articles that I might be able to use under gravel filtering and a combined airlift pump system. Do you see any problems with this setup? I plan to include skimmers to help keep the debris down.
    How heavily stocked is your 27 year old pond with fish and plants? Do you have any issues with planting media getting free in the pond and plugging the gravel filter?
    Thanks for taking the time to answer my questions!

  10. Mike White September 24, 2018 at 6:12 AM #

    Hi Dan,
    Your system should work just fine if the airlifts are setup correctly. My pond is the home of 75 koi. There are also some goldfish and golden Orfes. I also have frogs, trapdoor snails, and crayfish. I do not know if planting media would be a problem. My pond system is a system of two ponds with the water flowing from one pond into another pond and then back. The first pond has water lilies in it with planting media and the second pond has koi and under gravel filter. There are a lot of plants in the second pond but they are planted bare root in the rocks on the edge of the pond.

  11. Dan September 24, 2018 at 8:29 PM #

    Thanks for the response. Your ponds sound very impressive! That’s a lot of koi.
    So only your second pond (the koi pond) has an under gravel filter? Do the ponds have a secondary or backup filter system?
    Have you posted photos of your pond anywhere? I would very much like to see what it looks like.

  12. Mike White September 26, 2018 at 7:42 AM #

    Hi Dan,

    Yes the pond has a couple of other filters on it. The water from the undergravel filter flows to a lava rock filter that has 3,000 lbs of lava rock in it. There is also a homemade box filter with PVC ribbon and filter pads in it. There are no pictures of my pond on the internet.


  13. Adam October 26, 2018 at 1:23 AM #

    I’m new to building ponds and have read both your articles. The idea of a large pond with a low maintenance filtration system sounds like a dream come true, but I would like more details on how to calculate the size of the pump I would need for my pond. I would also like to know more about how to physically assemble this system. Can you point me towards some helpful resources?

  14. Mike White October 27, 2018 at 7:15 AM #

    Hi Adam,

    The size of the pump is more dependent on the volume of the pond to determine the smallest amount of flow that is needed. I use the following formula. Less than 1000 gallon pond at least a 2000 gph pump, 1000 to 4000 gallon pond a 4000 gph pump, 4000 to 10,000 gallon pond a 5000 gph pump and 10,000+ pond a bigger than 5000 gph pump.
    It is easiest to assembly right in the pond when building the pond. Pipes should be drilled ahead of time.


  15. Kirk December 30, 2018 at 4:20 AM #

    Hi Mike,
    I am constructing a raised 25 foot long by 5 foot wide by 2 foot deep (~1,800 gal) koi pond with a waterfall at one end and skimmer at the other. I would like to install an under gravel suction grid covering the entire pond bottom that connects to an external bead / UV filter with a 4,000 gpm pump. I was going to use six 25 foot long 2 inch pipes spaced about 1 foot apart and manifolded together at both ends. I was going to have a 6 inch deep bed of 3/4 -1 inch round gravel. Are the pipes of sufficient size and spacing or do you recommend a different configuration? Is it alright to connect the filter/pump to the grid at the skimmer end or should or do I need to connect in the middle of the long dimension of the grid for a more balanced suction? Is 6 inches of gravel too deep? I realize that this only leaves 18 inches of water but the 18 inch maximum depth is being dictated by local building codes. I will have a pretty heavy fish load. Would it be better to use a bigger pump? My preference is to substantially overdo the filtration.

    Thank you for your help.

    • Mike White February 8, 2019 at 5:16 PM #


      I am sorry I haven’t gotten back to you sooner. Five twenty four ft pipes will work just fine. Space so you have 6 inches from the side wall for the first one and then 12 inches on centers for the rest of them. I would think that everything should work just fine. You might want to go with a little bit larger pump depending on what external bead filter you are using. Some bead filters will cause some pretty high head pressures.


  16. Tony February 8, 2019 at 10:50 AM #

    I am looking to build an under gravel suction grid in my pond. I live in southern New Mexico and there is sand that blows around. I will avoid sand in my yard but I fear sand will be blown in either way. I am concerned that the pipes will clog with sand in the grid. Is there a way to avoid this or a way to design it so clearing the sand will be easier than digging up the whole system?

    Thank you in advance,

  17. Mike White February 8, 2019 at 5:05 PM #

    Hi Tony,

    I would not think that the sand would clog up the pipes as the pump would probably suck it in as the sand is much smaller than the holes. What I would see as possible problems would be that the sand could clog up the area between the rocks and give the system a problem with self adjusting flow. I think the big thing would be how much sand we are talking about and the size of the grains. I sorry I can’t give you a better answer but I have never seen this type of problem.


Leave a Reply