A few of you may have seen my summary posts on the RC thread and I wanted to re-post it here as well. It's still a work in progress, and I'm working on a "Part 2". I've also posted in on Reef2Reef and plan to post it on several other sites as well. So if anyone has comments, suggestions, or input of any kind, I gladly welcome them. Also feel free to post pictures of any kind of example (good, bad, ugly) that I can use in future summaries.

So here it is:

Algae Scrubber Build Basics Summary

There are many options when it comes to building your own scrubber. Hopefully this series of posts will wrap up all the basic concepts as clearly and concisely as possible. It seems that many people look at this concept and think they can make some kind of major improvement, and post their idea. Then, a newbie can look at that and think that it makes great sense and mimic it, without understanding the pitfalls. I feel that it is important to understand the concept and purpose behind the basics, and why you should rarely deviate from them without fully understanding the basic concepts first.

Most of the issues people encounter with their DIY builds can be directly attributed to missing one or two of the basic principles. So consider this a "best practice" summary. That is, there are several ways to build a scrubber, but there is always a cheap and easy way, as well as the most effective and efficient way.

The Vertical Waterfall Algae Scrubber

99% of this summary is dedicated to the basic of designing and constructing a modern, vertical screen, waterfall-style Algae Scrubber, illuminated from both sides. This is simply because this design is currently by far the most efficient and effective type of Algae Scrubber design, and the only one that I recommend building. Mention is made of what considerations to make for building a single-sided or horizontal scrubber.

The vertical waterfall style algae scrubber is very efficient for a couple of reasons. The design allows you to illuminate the algae from both sides and place the lights very close, which reduces the distance that light has to travel through water in order to reach the algae. The water flow over the screen is relatively thin and moves fast; this is important, because algae requires turbulent flow in order for nutrients to reach the cells on a microscopic level (this is referred to as the “boundary layer”). Because of both of these factors, vertical scrubbers require less light and material than the older style horizontal scrubbers for equivalent filtration capacity.

The Slot Pipe, Screen, and Fasteners

The principal component of the modern vertical waterfall Algae Scrubber is the slot pipe and plastic canvas screen. The screen is inserted into a section of horizontal PVC pipe which has a slot cut into it. The screen is typically held in place with a fastener through a cut-out section of the screen, like this:

Slot Pipe

The slot pipe itself should be Schedule 40 PVC at a minimum. Don't use vent pipe, flexible hose, or thin-walled PVC. The reason is that cutting a slot in the tube weakens it enough that it can bow over time and cause the slot to change width. This is especially true for a wide, tall scrubber. Anything over 24" should probably have 2 slots, with a small section in the center left uncut (maybe 1/4") with some kind of means of support at the center. It may look fine at first, but with heavy growth, you start to see this:

I'm not saying that it won't work, and I'm not critiquing the design pictured above. It just got me thinking that it can't be good to have that pipe flexing that much. So this is a recommendation I am making for scrubbers with screens over 24" wide. But check out that growth!!! That's 6.5 pounds of algae (drained).

Pipe Diameter

When running a scrubber fed directly from the overflow, there is usually no need to deviate from the size of the drain pipe that you currently use. But, that all depends on your total resultant flow to the pipe, length of screen, etc. In the majority of cases, this doesn’t change anything. However, if you had a large tank, and were combining 4 1” drain lines together to feed one large scrubber, you have different considerations.

If you're doing a pump fed scrubber, either top of tank or sump, try to match the size of the slot pipe and other PVC components to the recommendations from the pump manufacturer, and if in doubt, err on the side of larger pipe. Larger pipe puts less head pressure on the pump, which will increase the flow rate.

Cutting the Slot

The width of the slot should be approximately 1/8" wide, the same length as the screen which you are using (as exact as possible), and as straight as possible. Cutting a straight, even width slot is arguably the most difficult part of building a scrubber. It is also one of the most important steps. If the slot is crooked, the screen may not hang properly, and there may be areas where the flow is lighter and heavier. This can result in sections of weak growth.

However you go about this, you might need a few tries to get good at it. PVC is cheap, so practice on a small section so you get used to how this is done. It’s going to take a little time and patience – don’t rush it. Mark the slot, and cut carefully. If you mess up, do it again. Once you solvent weld an end cap and a union to it, it is not as cheap to replace (but still relatively cheap).

Doing it right means using a power tool. Probably the most common tool for this use is a rotary tool, or a Dremel, using a cutoff wheel:

Very recently, someone reported great success by using an oscillating tool. So if you’re looking for a reason to add another tool to the arsenal, here’s your chance.

Another idea is to use a router. Since plastics are ‘grabby’, and plunge routing plastic takes extra grip and control, extra precautions apply - so this is not for the inexperienced, but it makes very straight slots. This method requires building a guide jig for the router, and a clamping system to keep the PVC firmly in place.

Reportedly, the best way to cut a perfectly slot straight is by using a table saw. However, there really is no way to cut a slot in the middle of a section of pipe (or across the whole length of the pipe) without removing the anti-kickback device and lowering the pipe on to the spinning blade. You could also do it with a handheld circular saw, but like the router method, you would need a jig and clamp system, and you still are lowering a spinning blade onto the PVC (if you’re cutting the slot in the middle and leaving the ends uncut). I have never tried this and I do not recommend this method because it sounds too dangerous.

However you decide to cut the slot in the pipe, remember to observe all necessary safety precautions. Either that, or have someone else do it for you.

Screening Material

The search for the "perfect material" for growing algae in the vertical waterfall configuration has always come back to the same material, the Clear #7 Mesh Plastic Canvas. This can be found at almost any fabric (Jo-Ann Fabric) or craft-type hobby store (Michaels), usually in the knitting materials section. I have seen it at Wal-Mart, next to all the yarn. If you google “plastic canvas” you’ll find hundreds of links. You can feel free to experiment, but this material has been proven to work over and over again.

The reason that it works so well is because it is cheap, flexible, light, and easy to rough up. It is also translucent, which is critical. As algae grow thicker on a screen, the outer layer starts to block light to the lower layers. This causes weakening of the algae at the point of adhesion to the screen material. Allowing light to penetrate to the base of the screen from both sides helps to prevent the lower layers from weakening and detaching from the screen. Any material that blocks light is fundamentally inferior, all other things being equal. This also means don’t use any of the colored screening materials that you will see.

A brand new screen is slick and smooth. Except for the top edge of the screen, it needs to be roughed up with a hole saw or a rasp so that the algae has anchoring points. This is a critical step and it absolutely must be done properly. Once the screen is established, algae can grow very thick, it can hold a significant amount of water, and it will get heavy. I have seen pictures of screens that weigh several pounds. This weight is distributed across the screen, and with a rough base, it will hold very well.

To rough the screen up, you want to use a bi-metal hole saw. These saws have teeth that protrude out to a sharp point, and are great for tearing up the screen

As the above picture shows, you need to drag is back and forth across the screen. I suggest you sit in the garage or outside with a cutting board on your lap and lay the canvas on that. When you start roughing the screen, do a lot of random movements. Don’t concentrate on one spot for very long or you’ll wear through the screen and tear it. Rotate the screen often. You can also tilt the saw bit as you drag it so that the teeth won’t snag the screen. You should get a good heaping tablespoon of plastic from each side of a 10x10 screen.

After it’s all done, your screen should look like this:

Give the screen a good rinse and a light scrub with an old toothbrush to clear off all the pieces that are ‘hanging by a string’. These pieces are inert plastic, but still, you don’t really want these floating around in your aquarium.

Though I haven’t used one myself, I have heard that a rasp (available at any hardware store) is also great for roughing up this material. There are handheld rasps (similar to a Ped-Egg for your feet, except much sharper), and there are bit-mounted rasps (for power drills). It looks like it could be a little easier to tear the screen with this tool, but there is much less elbow grease involved. The drill bit rasps I have seen don’t appear to have protrusions large enough to “dig” into the screen and result in a cactus-rough surface.

Some have tried other methods of roughing up the screen. 80 grit sandpaper will definitely rough up the screen, but it doesn’t leave it ‘prickly’, so it’s not very effective. Recently, someone put a screen in the oven for a few minutes, then sprinkled it with aragonite sand and pressed it on; this blocks the light from penetrating to the base from both sides, but it’s still a cool idea, so time will tell on that one.

The bottom line here is that there really is no shortcut. You only have to rough up the screen once, so stick to the tried and true method, unless you feel like experimenting on your aquarium.

As previously mentioned, the top edge of the screen (where it is inserted into the slot pipe) should not be roughed up at all, or else algae can easily grow into the slot and restrict the flow. The amount of material you leave smooth depends on your design, mainly, how far you plan on inserting the screen into the slot pipe. Regardless of this measurement, the smooth section should extend at least 1/8” down below the slot.

I repeated get questioned about the risk of the slot clogging with algae and overflowing your tank. Instead of addressing this concern here, this will be addressed in Part 2 (which I again promise to get done soon).


Most people use zip-ties to secure the screen to the slot pipe, since they are cheap and easy to use. Generally you have to use new ones each time you clean the screen (because you have to cut them), or you can re-use them if you’re handy with the tip of a razor blade and can get them to release.

You can also use releasable zip ties (I found them in the electrical section of Home Depot), but they’re not as easy to release as you would think they would be, unfortunately. I use them, and I still need a mini-screwdriver to take them off.

Velcro straps can also be used, however they need to be free of adhesives, and I have never been able to get an answer from manufacturers about the stuff that has hooks on one side and loops on the other (such as OneWrap) regarding their safeness in aquaria. “Medical Grade” would be the safe way to go, but I haven’t looked much into those.

The latest one brought to my attention is the nylon beaded cable tie, which, like the zip-tie, is inert, but super-easy to release, so it’s good for months and possibly years.

The only drawback I can see it that you can’t cinch them down super-tight, but this doesn’t matter unless you hold the screen in place at the center only with the screen inserted all the way into the tube. Most people hold the screen in with one tie at each end, and maybe one in the middle, so this is less of an issue.

Another method of holding the screen in place is to use plastic shower curtain rings. It works very well, however you should have extras on hand in case one breaks when you’re removing it.

Yet another is to use sections of PVC pipe, one size larger that the slot tube, to make rotatable rings. This works very well, just make them narrow so that they don’t block too much flow.

I’ve seen a few designs that don’t use fasteners, but there are considerations to make with these designs. I decided to discuss these issues in Part 2.

Preventing water creep along the pipe

This is an important part of the slot pipe design. The issue here is that when you are running water though a horizontal pipe with a slot in the bottom of it, the water has a tendency to creep along the bottom edge of the pipe, not matter how perfectly horizontal you install it. If your scrubber is positioned completely over your sump, this may not be a big issue, as any errant flow will just drip into the sump. If your pipe ends extend close to the edge of the sump however, a steady trickle of water can result in gallons of water on the floor. I can attest to this.

The solution is very simple. Right at the point where the slot ends, place a large plumbing gasket, such as an o-ring (use several, they’re cheap), or even a bulkhead gasket. The point is that it needs to prevent water from getting past it. I can tell you from experience that a standard zip-tie will not work – the profile is too low.

In many cases, you need to put this piece on the pipe before you weld the cap on the end of the pipe, or else it will be very difficult to install. This one is a ballcock washer, and definitely needs to be installed before the cap is installed:

Planning your Scrubber

There are 2 basic ways of supplying flow to a scrubber: directly from the overflow, or from a dedicated pump.

The above diagram does not illustrate the top-of-tank scrubber, which would apply to someone running a sumpless system, however this is just a modification of the pump-driven scrubber with the pump in the display tank.

The very first step you need to do before buying, measuring, or sketching up anything, is to decide how you are going to supply your scrubber, and determine what your available flow rate is.

Available Flow

In any case, you need to measure the flow rate. Do this step. It is critical. Do not, I repeat, do not calculate the flow rate based on pump curves and head-feet of pressure. This may sound like a total pain in the behind, but just trust me on this one. Would you rather go through all the effort of building a scrubber, only to have problems and find out that you didn't have as much flow as you thought you did? Believe me, I've been there.

If you've been reading this thread, you will see that at some point I started making a big deal about this. The reason is that it is a big deal and I think many people don't realize that their pump does not pump at the rated flow, and in the majority of cases, it doesn't come close to the flow calculated by using a standard head-foot calculator program or table. So I have chosen to make it the #1 priority for scrubber design, hands down. You have to know your actual flow.

For a drain fed scrubber, fill a pitcher with the water entering the sump. You will probably need to rig up a temporary pipe or routing configuration so that you can fill the container. For a pump-fed scrubber, set up the pump in a sink filled with water to the same level as your pump will be submerged, and connect the tubing required to reach the height of the connection to the horizontal slot tube, so that you mimic as best as possible the actual conditions. Backpressure created by the slot/screen is negligible unless your flow rate significantly exceeds 35 GPH per inch of slot length.

Now that you've done all this, fill the container and record the time it takes to fill it. Do this at least a dozen times. The way I do this is by using a recording device, like a digital voice recorder, and just calling out "Go" and "Stop", then afterward, playing it back and using a stopwatch to get the time intervals. You could also have someone else run the stopwatch and write down the times. Average out the times and then figure out how many gallons per hour of flow you are actually getting. If you have multiple drains, measure and extrapolate GPH for each individually, and then add together.

For instance, if you are using a 1/2 gallon pitcher, and it takes 4.5 seconds to fill it, then you would have (0.5 gallons / 4.5 seconds) x (3600 seconds / 1 hour) which would be 400 GPH.

Don't be surprised if you have a lot less flow from your pump than you thought you had. I had less than 1/2 of what I thought it was. Head-feet calculations are usually way off, because most people don't use big enough return hose or have other restrictions in the plumbing. Some of it is inherent to reef-ready aquarium design (1" drain, 3/4" return, Danner Mag-Drive 9.5 and larger pumps need 1.5" return, see a problem?). So don't feel bad. A lot of people are in your situation, but they just don't know it.

Start with a clean pump. If your pump is not clean, soak it in vinegar for 15 minutes and scrub it good. After running a scrubber for about 4-5 months, your pump flow will drop about 15%, and by 6 months, it will have dropped by 25%, so you want to know your best-case flow and build around that. It's a lot easier to start with a throttled-back clean pump and open it a little when the flow rate decreases. Figure out your system flow rate, multiply by 80%, and that will be a good starting point. But, it's not going to kill you to start at full flow, and end up with a little less over time. You might just want to clean your pump a little more often, say every 3 months. So it's up to you. Just being aware of your system conditions puts you miles ahead.

Screen Size

Once you figure out your available flow, then it's time to figure out your optimal screen dimensions.

There are 2 ways of looking at this: square inches based on length and width dimension, and square inches based on illuminated surface area. The latter is technically more accurate, but since most people light both sides, the former is usually referenced.

For every gallon of water in your display tank, you need 2 total square inches of illuminated screen material. This means that if you run a screen that is vertical and lit from both sides, then you need a screen with dimensions (length times width) that is equal to the size of your tank, or 1 square inch of material per gallon. This is what you will see commonly referenced, and what I continue to reference for simplicity's sake. Double the dimensional measurement for a vertical screen, lit from only one side. Double it again for a horizontal or slanted screen.

Sizing of the screen generally does not require inclusion of the volume of water in the sump, unless you have some kind of bio-load in there, like a refugium for a recovering fish, pods, or a frag tank, etc.

So, just so we're 100% clear on this:

Vertical, lit from both sides: 1 square inch of screen material per gallon (2 square inches of illuminated screen area per gallon)

Vertical, lit from only one side: 2 square inches of screen material per gallon (which is also 2 square inches of illuminated screen area per gallon)

Horizontal: 4 square inches of screen material per gallon (4 square inches of illuminated screen material per gallon). Lighting must increase by a factor of 1.5 (discussed in the lighting section). Also note that this is a correction to what was listed on Post #1 of this thread (that post listed that a 10 x 10 screen was good for 40 gallons, instead of 25 gallons)

Screen Dimensions

So now that you know your actual flow rate AND the total size of your screen (and you need to know the total dimensional area for this part, not the total surface area), now you are ready to figure out your dimensions.

You want the flow to be as close to 35 GPH per inch of screen width as possible. You can get by with a lower flow rate, but your scrubber may not be strong enough depending on your bio-load. You can have higher flow also, which is generally not a problem as long as your screen is rough enough and you aren’t getting black slime algae (which is a sign of high nutrients, and needs more frequent cleaning until it lightens up. What you want to achieve is enough flow so that you have a full sheet of water across the screen off the bottom edge, like this:

Simply take your GPH (that you just measured) and divide by 35, and this will be your optimal screen width. Then, take that number and divide it into the gallon size of your display tank to obtain the height dimension of the screen. Use the size of your tank, not the gallons that you think are actually in it (so do not account for volume of Live Rock, fish, decor, etc) and do not include your sump volume (unless there is significant bio-load, as previously described). The result is the total area of roughed-up screen that you want.

In general, you want to add at least one inch to the height dimension for the section of smooth screen that will be inserted into the slot pipe. Specifically, you want to allow for the distance that the screen will be inserted into the slot pipe, plus at least 1/8" of smooth screen below the slot tube to help prevent algae growth into the slot. This "one inch" is just a good rule of thumb, and should be increased depending on the diameter of your slot pipe and how far you insert the screen into the slot pipe. A little extra smooth screen at the top never hurts, as it can be trimmed off later.

The critical area, and the only area that contributes to scrubbing power, is the roughed-up and illuminated portion of the screen. Figure out your necessary screen dimensions, then add the extra smooth section to the height dimension.


There are 2 basic types of light sources that 99% of people use: CFL and T5HO. In both cases, the optimum spectrum / light temp for growing algae is 2700K-3500K, with 2700K-3000K getting the best results.

Proper wattage of light and proper flow to the screen are the critical factors; color temperature / spectrum comes in behind those. You can use higher K ratings, but the real-world (anecdotal) evidence suggests that the optimal range for growing algae is heavy in the red spectrum. If you look at regular plant grow lights, you will find that most of them (especially LED grow lights) are very heavy in red.

Power Compact, or PC lamps, are not recommended, because they run way too hot for the amount of light you get out of them. I don't even care for them for tank lighting.

I have slightly expanded the discussion regarding LEDs. This is still a relatively unproved area, but as more people build LED scrubbers, more information is being confirmed. There are still plenty of unanswered questions, so LED scrubbers are really more of an advanced subject.

Quantity of lighting is dependent on the size of the system. In general, you want 1 watt (actual, NOT “incandescent equivalent”) of light per gallon for optimal scrubbing power. You can get away with less, but at the worst case you should use no less than 0.5 watts per gallon.

When I say “watt per gallon” here, I am referring to the size of the display tank, which is what is used to calculate the size of the screen. For the majority of people, the size of the screen matches the display tank size. If you decide to under or oversize the screen, you would then match the wattage required to the screen size. So if you have a 100 gallon tank, but you build your screen like it was on a 150 gallon tank, then you would want 150 watts of lighting.

As you will notice throughout this thread, it is generally stressed to follow the 1 watt per gallon guideline. This is because it solves many scrubber issues. The reason behind this is scientific. Light interacts with algae and causes N and P (and ammonia & nitrite, among other things) to be absorbed, and chlorophyll is created (among other things). The more light, the more nutrient reduction you get. There is a direct correlation between the quantity of light supplied and the amount of nutrient reduction capability.

A horizontal or slanted scrubber requires 1.5 watts per gallon, without exception. This is mainly because the horizontal scrubber has a much larger surface area to cover, so you need more light sources to spread the light out evenly. Also, horizontal/slanted scrubbers are generally not as efficient, unless you use a dump-bucket or surging style, which are more complicated.

You need to run your lighting for 18 hours on, 6 hours off. All life needs downtime. Plants are no exception. They have adapted to the environment over millions of years, and as the saying goes, you can’t fool Mother Nature. So don't go thinking that you can run lights 24/7 and get 25% more algae growth, it doesn’t work that way. The lights should be run on the reverse cycle of your display tank lighting; this assists in maintaining pH at night, as well as spreads the light-induced heat load more evenly throughout the day.

You want the lights as close as possible, within reason. The effective power/intensity of light follows the inverse square law. If you move a light twice as far away, the intensity drops by a factor of 4. If you move it twice as close, you get 4x the intensity. The balance point seems to be about 4" from the screen for CFL, and about 2" with T5HO. The reason for 4" away for CFL stems from hot spot issues due to the concentrated signature of the lamp; CFLs need to be a bit further away to cover the proper area without too much intensity. T5HOs do not have this problem, as the light is very evenly spread.

As far as spacing is concerned, CFLs need to be spaced according to the allowance of the design. If you need 2 per side, just position them for the best coverage. This is really on a case-by-case basis. As for T5HO, you generally want a lamp spacing of 2-3". For T5HO, your scrubber will generally need to be designed around the lamps and spacing. CFLs are more flexible in this respect, allowing a variety of configurations.

The lamps must be replaced every 3 months. This is not just a rule for scrubbers, you will see many people make this suggestion for refugium lighting as well. That is because there is a power drop-off and a spectrum shift that takes place over time, and when you go much past 3 months, you hit that drop-off point. We can't see the difference, but then again, we're not algae - it can. The result is that your screen will slow down growing and reduce filtration, which you do not want.

The light source needs to be positioned so that it is pointing directly at the screen material. Do not place the fixture so that it points parallel to the screen (from the ends or the top), place it so that directs the light toward the screen. Perfectly perpendicular is optimal, but if you have to point it at somewhat of an angle just to make it work, that will be fine. This is more of a concern for CFL than linear sources (T5HO), however I have seen a few T5HO build with the lamp 4 inches above the screen, shining straight down. They didn't work so well.

CFLs and Reflectors

The most common CFL used is the 23W Spiral. This is the actual wattage, not the equivalent wattage. There are a few different type of CFL lamps, and each one has a different ideal application. These are: spiral CFL, linear CFL, floodlight CFL.

A bare lamp will work without a reflector, but it will do the job much better with one.

Spiral CFL

The spiral CFL is definitely the most common type available and the most widely used. There are 2 ways to orient the lamp: with the end pointing at the screen or with the side facing the screen.

With the end lamp pointing at the screen, a reflector is a must-have piece of equipment. Without a reflector, a lamp pointed directly at the screen will do very little, since a small percentage of the light comes out of the end of the lamp. The cheapest, easiest, and most common reflector for this orientation is the dome-style reflector, which is available just about anywhere.

This reflector provides a wide light signature, since the side light is reflected to the front. They come in several sizes, and you want the biggest reflector you can fit in the space.

The dome reflector does the job just fine, and one of these should be used at a minimum. However, it is by no means the ‘perfect’ reflector. The reflector has a ribbed, dull surface that does a good job of diffusing the light, but it is not as efficient as a shiny, highly polished spectral reflector. Also, about 1/4 or so of the lamp (depending on the brand) sticks out past the reflector, and most of that light does not get directed toward the screen.

Using several different sized round reflectors can be done also when space is a limiting factor, even though the smaller reflectors are not as effective.

With the side of the lamp facing the screen, the reflector is usually a DIY job With a side-style orientation, more light is shed directly to the screen, but you still need to re-direct the light from the sides and back of the lamp towards the screen. There are a few fixtures that you can buy with integral reflectors, but most are very small. Most off-the-shelf light fixtures are for use in a shop or garage and have a half-round solid section, which may or may not have a reflector. If it does not, you can simply line it with aluminum foil or another highly reflective material.

The setup below uses off-the-shelf shop lights. The fixture on the left has reflective material installed.

The advantage to the side-lamp design is that since you can hang the lights from above, they generally take up less space (depending on your reflector) versus the dome reflectors.

In order to spread the light out evenly and wide (but not too wide or you’ll lose intensity at the screen), you want a wide reflector. Finding such a reflector is not easy. Since spiral CFLs can be considered a point source (more of a “blob” source, but this is for simplicity), using an HID reflector can be effective. A DIY beer can reflector can work also. Even some Mylar or aluminum foil will do the job.

Searching for a flexible reflector material and making your own reflector will yield the best results. A properly made reflector for a side-lamp orientation, such as the one shown below, can direct almost 100% of the light toward the screen.

Linear CFL

Linear CFL lamps are commonly referred to as twin, triple, or quad tube, etc. They are similar in nature to Power Compacts in that the lamp is in a “U” shape, but commonly have an integrated ballast like a CFL. They are usually higher wattage than standard CFLs, are more intense, and can run hotter. However, since they use the screw-in base just like CFLs, they are easy to use and I have recently looked at a couple of nice builds using them, so I thought it was worth adding a section covering them.

Linear CFLs would be installed similar to the sideways spiral CFL, hanging the lamp from above. Reflectors are generally the same principle; however the source is now more linear, so your reflector in turn should follow the line of the lamp and curve around it. Here are a few of the better ideas for this that I’ve seen. One uses cut-up linear fluorescent reflectors, the other uses mirrored acrylic.

A reflector similar to the last one in the spiral CFL section could be done. Because the lamp profile is more linear, the reflector would be slightly different dimensions – probably more square than rectangular.

Floodlight CFL

The floodlight CFL is simply a spiral CFL enclosed in a lamp housing like you would see for a standard incandescent floodlight. They are not very efficient at spreading light when placed in close proximity to the screen, as the light is diffused at the end of the ‘bulb’ and the reflector is of a small diameter. However, they are good for use on smaller, narrower screens – ones that have one dimension less than 6 inches. They should generally not be used for primary lighting, unless you are running a small scrubber, like for a 20-35 gallon tank

You can see in this picture that the floodlight only provides significant light to the area directly in front of the lamp – and that’s the only place that’s going to provide adequate filtration:

They can be useful in situations where space is highly restricted, but for larger scrubbers, more total wattage will likely be required over what would normally be needed.

They can also be helpful to supplement light from dome-reflector setups that just need a little more light but there’s not enough space for another dome.

One thing to remember when handling CFLs: install them gently. Most people are used to twisting in an incandescent lamp tightly. CFLs fracture easily at the base where the element (tube) meets up with the ballast. Cranking on them like causes these fractures. So if you can't grab on to the base to tighten, just get the lamp in there snug enough for the connection to be made. This goes for the lights in your house and office also - it's the #1 reason why CFLs burn out early.

T5HO lamps and Reflectors

T5HO lamps are inherently superior to CFLs. They spread the light out more evenly than CFLs, and they can be placed closer to the screen without overpowering the algae. Notice in this picture below how use of T5HO can result in almost perfectly even light coverage:

With that factor alone, bare T5HO lamps likely fall in between bare CFL and properly reflected CFL as far as scrubbing power is concerned.

Reflected T5HO is very arguably superior to all. However, it is more difficult to build a T5HO scrubber. You either need to buy and build around a stock fixture, like the Nova Extreme 1126/1127, or build something upon which to mount and protect endcaps, then connect to a ballast. Also, since the reflectors are generally one of the more expensive components, enclosing the screen in an acrylic or glass box is pretty much a given. However, in my opinion, it is worth every penny.

As far as lamps go, the same rules apply. I personally use the PlantMax 3000K Red/Bloom lamp, I can get them for $4 in packs of 8, which is about the cheapest I’ve been able to find for good quality lamps. Here’s my scrubber, Revision #2 (Rev 3 coming soon)

With a custom build, you can really use all the power of a T5HO lamp. The TEK-II reflectors from Sunlight Supply are expensive, but I have had great results. Other individual-lamp reflectors, such as Ice Caps, work very well. Stock fixtures work best when each light has an individual reflector, but just about any reflector will do better than none.

Sunlight Supply also makes a T5HO fixture that needs no external ballast, and you can daisy-chain up to 10 of them together. However, the reflector they make for it is just plain aluminum and nowhere near as reflective as the TEK-IIs or Ice Caps, which use Anolux-MIRO IV (which is something like 96% reflective)

The disadvantage to T5HO is that you’re pretty much locked into a dimension that you have to build around. T5HO lamps come in standard lengths of 24”, 36”, 48”, and 72”. There are shorter lamps, such as 18”, but it’s difficult to find lamps in the right color temperature, and they’re generally much more expensive than 24” or 48” lamps (which are the cheapest T5HO sizes). This means that the dimensions of your scrubber are locked into about 20-22” in length, which is the illuminated length of the lamp (the fixture is 24”, the lamp right around 21” long). This means your flow to the screen has to be around 700-800 GPH. You can make the screen narrower, but you will end up wasting part of the lamp. The benefits of T5HO over CFL would outweigh this loss to a certain point though.

This can be overcome if you have a dedicated fish room, or enough space to make a vertical T5HO scrubber, like this:

And to my knowledge, those endcaps are no longer available, or else I would totally be using them.


LEDs are a completely different source of light. Fluorescent, metal halide, HPS, and other HID lighting are all mercury based, and the light is shifted from the ultraviolet range into the visible range with phosphors. LEDs emit certain colors of light depending on the compounds used in the diode itself, so it is initially visible light; phosphors are then sometimes used to shift wavelength to achieve various colors.

Generally speaking, LED lighting has advantages over other type of lighting. The most obvious one is lamp life - they never actually burn out (unless you drive them too hard). LEDs have what is called an L70 (or L80) rating, which is the number of hours, running at rated junction temperature, at which the lumen output will have dropped to 70% of its original output. At this point in time, that is usually about 50,000 hours. If they are on 18 hours/day, that's about 7.6 years to L70. Some negative factors for LED are the long-term phase shift and the effect of steadily decreasing output. Phase shift is the reason that most small municipal airports are avoiding LED lighting: white LEDs are actually blue with phosphors added, and they ‘fade’ over time, and shift to blue. Runway lights are white, taxiway lights are blue, and getting them confused is bad. The LED industry is rapidly evolving, so the L70 numbers will continue to increase, cost will decrease, and issues like phase shifting will likely be improved upon.

But at the current state, there is a decent mathematical reasoning for going LED. Just using T5HO lamps as a baseline, let’s say you need 4 24W lamps ever 3 months for a 100 gallon scrubber, each costing $4 each. That’s $64 per year, and over 7 years, that’s $448. If you use 23W CFLs and you can find a stellar deal on them, you might be able to get them for $1 each, or about $112 over 7 years. So the cost of designing and building and LED fixture falls somewhere in between the two, IMO. If you DIY it, the material will cost you a couple hundred bucks, and the design time you’re going to spend either way, so most of that is offset.

The desired spectrum can theoretically be tuned to exactly what you need. This means that you can potentially overpower a scrubber if you use the same wattage as you would use for a CFL or T5HO scrubber, and just like you can bleach your corals with too much light, you can bleach algae with too much light. Also, it is important to note that there has been no study that I could find that indicates what exact LED spectrum is ‘perfect’ for algae growth for this specific purpose. At this point, you probably should not rely on an LED scrubber for total filtration, only for supplemental filtration. There are many unknowns, but it has great potential.

There’s not a lot of information regarding how these factors will impact algae growth. The phase shift is probably the biggest factor that is not known. Although, anyone who builds their own LED fixture will likely want to build another, better one in 2 years when more efficient and cheaper LEDs come out. So if you feel up to building your own LED scrubber, at this point, I say knock yourself out. Just make sure you do your homework before putting pen to paper; study other designs and learn from the mistakes and successes of others.

Just make sure you realize that 1) LED scrubbers do not have a long track record to speak of, 2) only recently have there been builds with any sort of success, 3) they are still not tested for long-term stability and reliability.

The general consensus among those that have built multiple LED scrubbers (believe me, there are not many that have built more than one) are that at a minimum, you should use Warm Whites, and ideally you want to use 630nm Reds, or a combination of the two. The jury is out on the need for any blue at all, and the effectiveness of 660nm Deep Red might be offset by the fact that the 630nm Red is more intense and therefore more efficient. Remember, flow and light intensity trumps spectrum.

LED builds to date vary in size, shape, number and arrangement of LEDs, use of lenses and/or diffusers, and on and on. This one uses 630nm Reds, warm whites, and in the middle are a couple of 660nm 5W Deep Reds, and a 5W Blue. This one uses bare LEDs and a diamond diffuser plate to spread the light.

This scrubber uses Warm Whites and Reds

This is an older one I found, it uses a panel covered with a combination of red and blue 1W LEDs. It was very expensive at the time ($700 I think)

There are also LED floodlights starting to hit the market. This one below is a 3000K 75W incandescent equivalent, which equates roughly to an 18W CFL:

I like that this one uses high-quality Cree LEDs. However, since there is very little data on the effectiveness of these lamps, the same guidelines for CFL floodlights apply to these. Spreading the light out adequately is the main problem with LEDs in general. It’s difficult to tell from this picture, but it appears that the light is focused similarly to a CFL floodlight:

An off-the-shelf LED fixture is generally more expensive than a DIY job. For scrubbers, there are very few choices in LED off-the-shelf, it’s pretty much limited to plant growth lamps. There’s plenty of them out there, but they’re rather expensive and lots of them are made overseas, where they use cheap LEDs, under-drive them so they don’t blow, and put them in cheap housings that aren’t made for use anywhere near water. The American made fixture are much more reliably built, but the prices are coming down. It’s only a matter of time before LEDs are tried and tested.

Electrical Protection

This should really go without saying, but you should always plug your lights into Ground-Fault protected receptacles (GFCI). You should actually have all your equipment plugged into GFCI receptacles, and on as many separate circuits as feasible – but that’s another discussion.

Always use waterproof sockets for your CFLs and end caps for your T5HO lamps. This is a little more expensive, but is necessary to avoid corrosion and electrocution. Generally, waterproof CFL sockets do a pretty good job of sealing the base and socket from moisture, but they still should be silicone sealed for an extra layer of protection. T5HO waterproof end caps do an excellent job of sealing the end of the lamp, but the wires that feed into the bottom of the sockets are not sealed, so after all wiring is complete, you need to fill in the bottom with silicone caulk.

You can't see it, but there will be tiny amounts of salt spray that will build up where you screw a CFL bulb in, and also where you make electrical connections. When the buildup gets thick enough, it can short out and trip a breaker or GFCI receptacle, or shock you. So each time you replace a CFL screw-in lamp, re-seal it. You should be able to pour water over it without it causing a problem (but don’t try it). Use GE Silicone I Door & Window caulk, which is generally accepted as aquarium safe, especially since you don’t intend for it to be in direct contact with water anyways.

Spray Protection

As far as spray and salt creep is concerned, you want to avoid buildup on the lamp itself. No matter what you do, there will still likely be some buildup due to evaporation, so you will want to wipe off the lamp periodically (as needed). Make sure the lamp is completely cooled down before wiping the lamp off, remove the lamp if possible, and wipe it down with a soft cloth and warm to hot water. This is rather easy to do in place with T5HO lamps, and more difficult for spiral and multi-tube CFLs; just be gentle so you don’t crack the tube.

The water will cascade down the screen smoothly, and then drop off the bottom edge. If you don’t let the bottom of the screen sit in the water, it’s going to pour off and crash down, and splash everywhere. That’s not the problem I’m discussing here, I’m talking about spray from the slot pipe or screen.

The ideal solution is to block the source of salt creep or spray – the slot pipe itself. While 99.99% of the water that cascades down the screen will stay on the screen, occasionally there will be droplets of water that pop and fly around, and over time these can cause salt creep. If you have a lot of spray on a consistent basis, then you have a different problem – this generally should not happen, and will be covered under the troubleshooting section in Part 2.

The best spray blocker is a box that totally encloses the screen on all 4 sides and bottom, and has a removable lid. Such a box would have a drain hole in the bottom, and would typically be made out of acrylic, but could be made from glass also. This is beneficial for other reasons too, but we’re just talking spray blocking right now. If you go this route, you want to read the section regarding enclosed boxes in Part 2 for more details - which is coming, I promise!!

Closing the top with a lid, or at least extending the blocker up to the top of the slot pipe, will minimize the random drops that occasionally fly upward, as well as evaporation. A lid should not lip over the outside of the box, rather the inside so that any condensation will tend to stay inside the box.

The next best would be an enclosure with an open bottom. The advantage to an open-bottom enclosure is that it's easy to build. If you extend the screen to the water level inside the sump, you virtually eliminate noise and microbubbles. Pictures illustrate this best:

The most basic spray blocker would just be a couple of plastic panels draped over the tube. Here are a few examples:

I don’t recommend enclosing CFLs in glass jars. IMO, it is difficult to allow for adequate air convection with such an arrangement. I don’t recommend using plastic bottles either. CFLs get hot; you can’t touch them when they’re on. If the lamp gets in contact with the plastic, it could melt.

End of Summary #1

As mentioned at the start of this summary, I will write up “Part 2” which will cover various troubleshooting pointers, enclosed boxes, growth types, customizing to your specific system, overflow protection, increasing flow, operation, maintenance, cleaning, etc

If you have any suggestions or corrections, please post them or shoot me a PM and I will note them for future summary posts.