Skimmer Design 101
Version 1.0, revised 09/26/2001
Shane Graber (aka
Over the last two years I have been looking through the various DIY skimmer designs published on the web like the ones that a person might find on various sites like http://www.ozreef.org/, etc. In seeing my first skimmer at the LFS, I thought "making one of those can't be *that* hard, can it?" Ever since then, I have been interested in learning about skimmers and their proper design. There are many good designs published on the web but the things that I *always* found lacking was:
I was confused by all of these questions until I purchased Aquatic Systems Engineering: Devices and How They Function by P.R. Escobal. This book effectively explained answers to many of my questions and it also delved deep into proper skimmer design and function in addition to many other subjects such as plumbing, flow, etc, for the aquarium. Anyone interested in DIY'ing anything should get their hands on this book as I have found it to be an indispensible book for the serious DIY'er.
First off, I would like to introduce a schematic of a counter current skimmer. In this case, I will illustrate skimmer function using an airstone driven counter-current skimmer design:
Water is pumped into the skimmer at the top of the reaction chamber via a pump that is sized and the flow rate is adjusted for the particular application. At the bottom of the reaction chamber, an airstone hooked to an external air pump produces millions of tiny bubbles. These bubbles begin to rise up through the water in the reaction chamber of the skimmer where they contact the water flowing down through the skimmer. During this contact time, the proteins and other dissolved organics that can be skimmed out of the water column are be attracted to the air bubbles due to the molecules' specific hydrophilic ("water loving") / hydrophobic ("water hating") nature. The attached molecules are then carried to the neck of the protein skimmer attached to the air bubbles. The "cleaned" water then exits the bottom of the skimmer which runs either back to the tank or to the sump. As more and more skimmed organics accumulate at the top of the reaction chamber, they are eventually pushed out of the skimmer through the neck and into the collection cup. From here the "skimmate" is drained off into a collection container for later disposal.
Next, I would like to introduce what the proteins and other organic molecules are doing when they come into contact with an air bubble:
The circle in the above diagram is an individual air bubble released by the airstones inside the protein skimmer. The smaller the air bubble the better as the smaller the air bubble per unit volume of air injected, the more water/air surface area is produced. Think of this as two identical containers: on containing a given mass of pea gravel. An identical container holds an identical weight of playsand. Which container has more surface area of the particles inside it: the pea gravel or play sand? The correct answer is the play sand. As the bubble travels up through the water column, the elecrically charged protein molecules (which contain elecrically polar and electrically nonpolar regions) are attracted to the air/water interface of the bubble. The polar regions of the molecule (made up of nitrogens, oxygens, etc) are attracted to the air/water interface and these polar "tails" stick out away from the air bubble into the water column. The nonpolar regions stick out into the air bubble because it does not "like" to be in contact with the polar solvent (i.e. water). If you could look at this bubble under high enough magnification down to the molecular level, the entire air bubble would look like a fuzzy ball with protein tails and other electrically charged tails sticking out from the surface of the air bubble. The polar regions outside of the air bubble stabilize the air bubble very much like a soap bubble in your kitchen sink or your washing machine. This is why a foam begins to build up at the surface of the skimmer. As the protein laden bubble reaches the top of the protein skimmer, the proteins begin to accumulate which creates a stable foam bubble. These stable foam bubbles take a long time to pop. Thus, the proteins slowly are concentrated at the top of the skimmer where they are slowly pushed through the "throat" of the protein skimmer and into the collection cup.
Different Skimmer Designs:
With all the different skimmer options out there, it's sometimes hard for the person new to the hobby to understand all of the "lingo" when it comes to skimmers. Basically, there's two main areas that are of focus when people talk about their skimmers: water flow and air injection method.
With water flow, there's two general methods:
These two terms simply explain which way the water is flowing with respect to the flow of the bubbles in the skimmer. A co-current skimmer has the water flowing in the same direction as the bubbles and a counter current skimmer has the water flowing against the upward flow of the bubbles.
For air introduction, there's many methods. Four of the primary methods are:
The air pump / airstone method is pretty straight forward: an airstone attached to a good air pump is situated at the bottom of the reaction chamber and water flows from bottom-to-top or top-to-bottom (co-current vs. counter-current) as pictured above. A venturi skimmer uses a venturi device to suck air into a water line as water travels thru a venturi device. A couple good pictures and websites on venturi skimmers can be found at http://w3page.com/fishline/gif/ventskim.gif, http://mars.reefkeepers.net/Articles/SuperSkimmer.html. A downdraft skimmer utilizes a pressure rated pump to push a high volume of water into a tall tube filled with bioballs. This tube has a valve attached that allows air into the mixing chamber where the water impacts with the bioballs in the tube. Geo has a couple nice shots of his DIY downdraft skimmer: http://www.homestead.com/geosreef/downdraftskimmer.html and Peter Z has a very nice cross sectional schematic of a downdraft skimmer: http://w3page.com/fishline/gif/ddskimpz.jpg. A beckett skimmer works similar to a downdraft skimmer except it uses a beckett pond injector and a pressure rated pump to make a great deal of foam. A couple good URL's to look at for how specifically a beckett is setup is: http://www.eparc.com/diy/skimmers/ian/hsa.shtml, http://nucalf.physics.fsu.edu/pfohl/Fish/Diy/hsa.gif.
By far the most efficient way to run a skimmer is in the counter current flow condition as it causes the bubbles to have a longer contact time with the water compared to the co-current flow condition. It was estimated by Escobal that some proteins take upwards of 2 minutes contact time with air to attach properly. Now when it comes to air introduction method, methods #2, #3, and #4 all use the pump powering the skimmer to introduce air into the skimmer at the same time the pump pumps water into the reaction chamber. While this works just fine, it does limit the user to not being able to independently control water flow rate through the skimmer and the air volume injected into the skimmer reaction chamber. With options #2-#4, if a person were to decrease the water flow through the skimmer the person would also decrease the amount of air injected into the reaction chamber. Option #1, however, is uniquely different from the other 3 methods as it *does* allow independent control over water flow through the skimmer and air flow rate into the skimmer. Because of this, counter current air driven skimmers are among some of the most efficient skimmers out there when designed properly. They are also some of the most energy efficient skimmers available. Many of the larger venturi, downdraft, and beckett skimmers require large energy hogging pumps to produce enough flow to suck air into the skimmer reaction chamber.
For the remainder of this tutorial, I will be referring to a counter current air driven skimmer. For all practical purposes, the following items are critical in skimmer design:
Let's attack each of the above points one at a time.
1. Water Flow Rate Through the Skimmer:
For optimum skimming using a counter current design (or actually any other design for that matter), the water in your tank should not flow through the skimmer any more than two times per day. For venturi's, downdraft's, and beckett's, this requirement is almost a virtual impossibility as the pumps need to pump the huge amounts of water to yield a sufficient quantity of air to skim. This may seem strange and even shocking to many people as they are used to the "Tim Taylor" - type skimmers (more power! Oh Oh OH OH oh!) that push huge volumes of water through a skimmer every hour. Many people believe that to skim more effectively and efficiently, you have to increase the water flow through their skimmer as the air bubbles will contact the air bubbles more times per hour. This is NOT the case. The water running through the skimmer is not the limiting factor when it comes to nutrient export. It's the amount of AIR that contacts each "drop" of water that is the limiting factor to how much a skimmer pulls out of the water column. Some organics require up to 2 full minutes of contact time with air bubbles in a skimmer before they are removed via foam fractioning. Thus the need for a slow wate flow through the skimmer is crutial for proper design and function.
According to Escobal, the following chart should be used to determine the flow rate for water through the skimmer (assuming a 2x per day turnover rate):
Find the net gallons of water contained in your system on the above chart on the X-Axis and follow that value vertically up the chart until you contact the green line. At that point, follow the line over to the left to the Y-axis. This is the flowrate that should flow through your skimmer for optimum skimming. This particular flow rate will run all of your tank water through the skimmer two times each day.
For example, let's say you have 100 gallons of water in your reeftank and you want to find out how many gallons per hour you should flow water through your counter current skimmer for optimum skimming. Find 100 gallons on the X-axis, follow the line on the chart vertically until you hit the green line. Now follow that over to the Y-axis to find out how much water should be run through your skimmer every hour. In this case it is 75 gph. Thus, for your counter current skimmer the tank water should not flow through the skimmer any faster or any slower than 75 gph. Pretty straight forward.
Using the above example for a moment, this should then raise another question: I'm flowing 75 gph through my skimmer. According to that, I should be running all of my tank water through the skimmer every 1.333 hours (100 gal/75 gal/hr) and not every 12 hours. Contrary to popular thought, this is not the case. Every time you run water through the skimmer, it dumps the water back into the tank of "unskimmed" water thereby diluting it. As water is continually pulled through the skimmer, it will then pull in "skimmed" and "unskimmed" water. As more "skimmed" water is pumped back into the tank, more and more already "skimmed" water will flow back through the skimmer. According to Escobal, the following equation is used to calculate when 99.99% of the water has flowed through the skimmer:
T = 9.2 (G/F)
Where T=time, G=total gallons of tank water, and F=flow rate (gph). The 9.2 is a purity coefficient that when used in this equation yields a 99.99% purity. For further explanation of this, please refer to Escobal's book.
2. Height of the Skimmer
This one's pretty straight forward. Make the skimmer reaction chamber as tall as you can to maximize the total contact time that the water has with the air in the skimmer. Escobal recommends at least 4 feet to 5 feet tall for the reaction chamber to optimize contact time but a person could go higher or lower than these recommended lengths if they so desire to fit their application. Fossa and Nilsen in The Modern Coral Reef Aquarium volume #1, page 259 recommend that the skimmer reaction chamber should not be any shorter than 28 inches to 32 inches "except for very small aquaria" so use your best judgement when designing and making your skimmer.
3. Amount of Air Pumped into Reaction Chamber
According to Escobal, the upper limit of the amount of air able to be inside the skimmer at any one time is 13% of the water volume inside the skimmer. Escobal does take the reader through the calculations on why this is the uppere limit, but I don't plan to as this calculation is quite drawn out. However, if you stop to think about it there can only be so much air injected into the skimmer before you overpower the water itself and the skimmer runs dry. This value just happens to be 13%. According to Escobal: "...the volume of air in a skimmer can never be greater than about 13% of the volume of the skimmer or the bubbles will merge. Experimental measurements are in agreement..." (Escobal, p. 96). So if you hear any manufacturers out there claiming that their product will inject more than 13% air into the skimmer body, then I would highly suspect that they're not going to be able to acheive the proper bubble size inside their skimmer.
There are two keys to injecting air into the skimmer for optimum nutrient export:
The first requirement is pretty straight forward. A small bubble size is desired as the smaller the bubble is the more surface area it has to the total volume of air. Let me give an example. Let's say we have a basketball. This basketball holds quite a bit of air inside of it. Now let's take the amount of air inside the basketball and put it inside a bunch of tennis balls. It takes a LOT of tennis balls to hold the volume of that one basketball. Now if we look at these two items, which one has more surface area? The tennis balls have more combined surface area than the one basketball. We can extrapolate this down to very fine bubbles. The smaller the bubble, the more surface area it has relative to the total volume of air. The more surface area the bubble has, the more dissolved organics are going to be picked up out of the water column inside the skimmer.
So now we know that we need to have a very fine bubble size. How do we acheive that inside of the counter current skimmer? The key is to use a very fine pore size airstone. The best airstones are made out of limewood but basswood and oak are also used as well. Oak airstones take a much more hefty pump than is typically used, but if used it does produce a very small bubble. For all practical purposes however, limewood and basswood are the primary airstones used for counter current skimmers. It should also be noted that only a certain amount of air can be pumped through a single airstone before it overpowers the airstone. Once the air into the airstone overpowers the airstone, larger bubbles will be produced. The key here is to use more than one airstone if necessary to acheive the correct bubble size. For small diameter counter current skimmers maybe one airstone is all that's necessary. For the much larger skimmers upwards of 4 to 6 or even 10 airstones may be required. It all depends upon the application. A person can also buy longer airstones if they so choose. The key is that you want a lot of surface area for the bubbles to form on so you can either go with may small airstones or one long airstone or a combination. You can either buy your airstones or you can make them. Buying them can get somewhat costly if you have a large skimmer as they can cost upwards of $1 to $3 per airstone depending upon the size required. Another option is to make your own airstones. This is much more economical in the long run. There are many good websites out there that explain how to make your own airstones.
The second requirement, that of minimizing turbulence of the air bubbles inside the skimmer, is also crutial. The reason for this is because of the different types of dissolved organics in the water column. Different organics are attracted to the surface of the air bubbles to a different degree depending upon the molecular composition of the dissolved compound. Some compounds are highly attracted to the air bubble. Others are only weakly attracted. If the air/water is turbulent inside the skimmer, the strongly attracted compounds will stay attached to the air bubble but the weakly attracted compounds will not and they will not be skimmed out of the water column. By minimizing the turbulence inside the skimmer, both the strongly and weakly attracted compounds stay attached to the air bubble and are skimmed out of the water column. What you want to see is a nice blooming flow of air bubbles upward in the water column. If you see excessive turbulence inside of the skimmer then the skimmer is not tuned properly.
4. Diameter of the Skimmer
This selection is also pretty straight forward. The larger the amount of water to skim, the wider the diameter of the skimmer needs to be in order to skim effectively. This requirement also ties into the previous section with the amount of air pumped into the reaction chamber of the skimmer and minimizing turbulence of that air flowing upward. Based on these requirements, Escobal published a chart detailing the specific diameters required for tanks of known volume. Below is the chart:
|It's pretty straight forward: find the number of gallons for your tank on the X-axis and follow that up vertically until you intersect the "12 Hour Limit Line" line at the top of the graph. Now look for the diagonal diameter lines and see which one comes closest to the line you followed up the graph. This is the diameter for the skimmer that you are designing. If the point at which you intersect the "12 Hour Limit Line" is between two of the diagonal lines you can either choose the diameter larger or smaller. Going larger will allow you more skimming ability. Going smaller should not seriously hamper your skimming ability. Or you can go with more than one skimmer. However, if you do go with more than one skimmer, their total cross-sectional area needs to match pretty closely the cross-sectional area of the single skimmer diameter that is required for the project.|
Now that the above has been explained, let's use a practical example:
Let's say you have a tank with 125 gallons of water in it and you want to design a skimmer for it. >From Section #1 above, you know that for a 125 gallon water volume you need to flow water through the skimmer at approximately 100 gph. From Section #2 above, you know that you need to make the skimmer reaction chamber as tall as possible. If the skimmer is to be sitting beside the tank you can make the skimmer reaction chamber as tall up to 4 feet or 5 feet tall if possible or even taller if you so choose. If it is to be an in-sump design you could go with a shorter skimmer as long as the reaction chamber is no less than 28 inches. Now that the height of the reaction chamber has been decided upon, you need to decide on the length of the airstones and the number of airstones to be used for the application. Personally for the 4" diameter skimmer design I like to use 4 airstones with a length of each airstone between 4" to 6". I would try to fit as many airstones as you can inside the cross-sectional area of the skimmer as possible to maximize airstone surface area which should make the finest bubbles. From Section #4, inspecting the chart for a 125 gallon water volume and a turnover time of 12 hours you should pick a skimmer with a diameter of 5". Since 5" diameter PVC is somewhat hard to find, another option you could go with would be to use two 4" diameter skimmers or three 3" diameter skimmers. These multiple skimmers would be connected in series so that water would flow through one and directly into the next one.
That's pretty much it. Use your new-found information to design a properly functioning counter current skimmer! :)