Skimming Principles

[tinygiants]

I am in the process of designing a skimmer for my tank. I found a few principals (Aquatic Systems Engineering by Peter Escobal) to base my design on.

Ideal skimming would be 2 - 3 tank volumes a day.
The formula for this is:
(total tank and sump volume / feed rate gph) X 9.2 (this is purity coefficent that will yield 99.9% processed water) = hours till processed

The goal of 2 - 3 times a day means we need to process every 8 - 12 hours.

In my tank it works like this:

( 150 g / flow rate ) 9.2 = 12
150 g / flow rate = 12/9.2
150 g / flow rate = 1.30
150 g = 1.30 flow rate
150 g / 1.30 = flow rate
115 gph = flow rate

This means by feeding my skimmer 115 gph, I will 99.9 % process my tank every 12 hours.

Dwell time is another issue. It is based on this formula:
skimmer volume / gpm = dwell time for water

Dwell time alone is not enough. We also want the air bubble dwell time. With this info we can figure bombardment rate

I have no idea how to get an accurate measure of a bubbles actual dwell time. So I pulled a number out of thin air of 15 seconds in 24" counter current design.

In my tank:
Using a 2 gpm feed into 11 g skimmer (12 dia X 24 tall) I get a dwell time of 5.5 minutes.
5.5 minutes = 330 seconds
bombardment rate = 330/15 = 22.

Bombardment rate = water dwell in seconds / air dwell in seconds
A perfect goal is considered to be 10.

I am way off of the goal, but is it bad?

Thanks for looking.
Dale

[Puffers]

While I am no skimmer expert trying to pull a number out of "thin-air" and apply it in a formula and then questioning the answer of it to say it good or bad is tuff. There may be info on the web in regards to a bubble buoyancy, (taking size into consideration) in which a similar sized bubble, flowing up a certain height will only travel so far, so fast. We could time the skimmer to see how long it takes to full up the riser neck with bubbles. Do this several times, take an average and see what the number may be.

Even with an accurate number there are other factors to consider. Bio-load, light heavy or normal? Do you plan on skimming wet or dry? How often do you plan on cleaning your skimmer? Then you have the real world "factor" versus a "paper" answer. IMO upgrade your skimmer by a factor of 2, both gph and height. In most reefs with skimmers it is probably the most important filtration device beside LR.

[Steven Pro]

You might be able to measure the dwell time by placing the skimmer in a small sterile aquarium setup as close as you can to the normal operating conditions. Then add food coloring near the feed pump's intake and time how long before the coloring comes out of the skimmer body.

[Puffers]

I thought of that but the water that comes out of the skimmer is not necessarily the dwell time I understand to be of importance. I may be wrong, if so be kind...

Isn't the dwell time considered the amount of time the bubble is inside the skimmer to the riser neck? So once the bubble is created at the impeller (assuming venturi driven) would be the start time of the watch and the end time would be when the riser neck "top" is or has the first bubbles?

[Steven Pro]

I don't know. I thought dwell time would refer to the amount of time water would potnetial skimmable material would be exposed to the air bubbles.

[tinygiants]

Water dwell is easy enough to figure, it is based on volume / gph.

You are right a number out thin air is weak. But I am sure it will be < 33 seconds to travel 24" in a 2 gpm environment.

My real question is in regards to the base "rule". How important is it? Which rules are better to break?

If I increase flow t 3 gpm, my dwell time is 3.66 minutes or 220 seconds. That makes my theoretical 15 sec air dwell = a bombardment of 15. Better bombardment rate, but I am sacrificing turnover rate. Which is more important? Where can I compromise?

[Puffers]

Most skimmers sold today don't talk about dwell time. In most cases, not all, they rate the skimmer by gph, ie turnover.

In this case it sounds like you want the perfect skimmer but it looks like you have to sacrifice one, you choose (gph) or(dwell)?

In regards to dwell being easy to figure please post that formula as I would like to input the results for my CV220 skimmer.

[barryhc]

Originally posted by tinygiants
Water dwell is easy enough to figure, it is based on volume / gph.

You are right a number out thin air is weak. But I am sure it will be < 33 seconds to travel 24" in a 2 gpm environment.

My real question is in regards to the base "rule". How important is it? Which rules are better to break?

If I increase flow t 3 gpm, my dwell time is 3.66 minutes or 220 seconds. That makes my theoretical 15 sec air dwell = a bombardment of 15. Better bombardment rate, but I am sacrificing turnover rate. Which is more important? Where can I compromise?

Two stage( or recirculating ) type skimmers are able to increase the "bubble dwell time" by many "factors" of magnitude. Including a "recirculation path", would allow you to tune your bombardment rate to "match" whatever flow you were trying out at the time.

In other words, you can build the skimmer, and then you can adjust it's bombardment rate, and "effective size" to some degree, after you have built it.

You are experimenting, this helps you experiment.

> barryhc

[H20ENG]

A roundabout way to check the bubble dwell time may be to count how long it takes for the bubbles to dissipate after turning off the pump.

[tinygiants]

Originally posted by H20ENG
A roundabout way to check the bubble dwell time may be to count how long it takes for the bubbles to dissipate after turning off the pump.

Of all the thoughts I have had about bubble dwell, this sounds the most promising.

Thanks for the idea.
Dale

[tinygiants]

Originally posted by Puffers
In regards to dwell being easy to figure please post that formula as I would like to input the results for my CV220 skimmer.

Water dwell time = vessel volume / feed rate

In my case:

Water Dwell = 11 G / 2GPM
Water Dwell = 5.5 Minutes

The first formula in the post deals with processing time when using recirculation method (returning processed water to the unprocessed source volume)

(total tank and sump volume / feed rate gph) X 9.2 [this is purity coefficent that will yield 99.9% processed water] = hours till processed

Dale

[jimbo045]

I say skimming is a must.

[barryhc]

I have done a good bit of surfing regarding skimmers and skimming today, and came across several interesting points.

Maximum saturation of bubbles in water is 13%. More than this causes the "bubble column" to collapse.

Some minimum height requirements that are probably only valid for nonrecirculating types.

More than one source claims that "turbulence" should be kept to a minimum. It had to do with "high" and "low" receptor "affinities" for various compounds or nutrients.

I didn't quite understand it, but, my previous "promotion" of recirculating type "skimmimg" may have been brought into question here regarding "turbulence".

> barryhc

[tinygiants]

I am leaning heavily toward a air stone skimmer for the turbulence reason. In order to use a beckett, the pump feed rate must be very high. In order to get a decent dwell time, you are forced into a recirc style. From a design stand point, I believe that the standard counter current air skimmer may be the most efficient.

[moumda]

Since you will have a set amount of water flow set by your pump (once you set your dwell time) couldn't you set skimmer performance with the air flow? That's how I control my skimmers. The more air the wetter the skimmate. JMO.

[Puffers]

I believe that the standard counter current air skimmer may be the most efficient.

First time I've heard that in a while. I've always thought my air driven cc skimmer seem to work better?

Is the reason for keeping the turbulence to a minimum have to do with the proteins actually sticking to the bubbles long enough to be processed (bubble pop) out?

[barryhc]

Originally posted by Puffers
First time I've heard that in a while. I've always thought my air driven cc skimmer seem to work better?

Is the reason for keeping the turbulence to a minimum have to do with the proteins actually sticking to the bubbles long enough to be processed (bubble pop) out?

Not exactly, It's the relatively low "receptor affinity" of some compounds that may not attach at all, if turbulence is too high.

If I find the link on that again, I will post it here.

> barryhc

[wrasselover]

Originally posted by tinygiants
I am in the process of designing a skimmer for my tank. I found a few principals (Aquatic Systems Engineering by Peter Escobal) to base my design on.

Ideal skimming would be 2 - 3 tank volumes a day.
The formula for this is:
(total tank and sump volume / feed rate gph) X 9.2 (this is purity coefficent that will yield 99.9% processed water) = hours till processed

The goal of 2 - 3 times a day means we need to process every 8 - 12 hours.

In my tank it works like this:

( 150 g / flow rate ) 9.2 = 12
150 g / flow rate = 12/9.2
150 g / flow rate = 1.30
150 g = 1.30 flow rate
150 g / 1.30 = flow rate
115 gph = flow rate

This means by feeding my skimmer 115 gph, I will 99.9 % process my tank every 12 hours.

Dwell time is another issue. It is based on this formula:
skimmer volume / gpm = dwell time for water

Dwell time alone is not enough. We also want the air bubble dwell time. With this info we can figure bombardment rate

I have no idea how to get an accurate measure of a bubbles actual dwell time. So I pulled a number out of thin air of 15 seconds in 24" counter current design.

In my tank:
Using a 2 gpm feed into 11 g skimmer (12 dia X 24 tall) I get a dwell time of 5.5 minutes.
5.5 minutes = 330 seconds
bombardment rate = 330/15 = 22.

Bombardment rate = water dwell in seconds / air dwell in seconds
A perfect goal is considered to be 10.

I am way off of the goal, but is it bad?

Thanks for looking.
Dale

I think your should look ratio numbers, like dwell, as a minimum recommended, not a basis for optimum design. Your 22 bombardment rate is better than designing for a 10. It is difficult to over size a skimmer....it is easy to undersize. Just make the biggest baddest one that you can fit in and afford. This is another area where larger is better......

[barryhc]

Originally posted by wrasselover
I think your should look ratio numbers, like dwell, as a minimum recommended, not a basis for optimum design. Your 22 bombardment rate is better than designing for a 10. It is difficult to over size a skimmer....it is easy to undersize. Just make the biggest baddest one that you can fit in and afford. This is another area where larger is better......

SO, "more is better" heh? You need to go out and buy a 7,000 HP car, TOMMOROW!!!!!!! (( better yet, this afternoon )

1 + 1 = 2, and the "inverse" of 85% is 117.6%, so how about explaining Chloramine to us again, "Wrasselover"?

With all my prayers! > barryhc

[barryhc]

Originally posted by barryhc
I have done a good bit of surfing regarding skimmers and skimming today, and came across several interesting points.

Maximum saturation of bubbles in water is 13%. More than this causes the "bubble column" to collapse. > barryhc

This "bubble collapse" would be in the Primary ( dwell ) chamber, of course, and not the "collection chamber".

> barryhc

[tinygiants]

Where did you find the 13%? The info I found quoted 20%.

I am still in design phase, so it is all just theory. I have not yet bought the air pump.
The PVC is just around the block, I may go get it this weekend.

[barryhc]

Originally posted by tinygiants
Where did you find the 13%? The info I found quoted 20%.

I am still in design phase, so it is all just theory. I have not yet bought the air pump.
The PVC is just around the block, I may go get it this weekend.

I think you are more up on this than I am "tinygiants", but again this is the ratio of oxygen to water in the "dwell chamber".

I think that I may have actually found this on a link within the links that you have provided here, I am not sure.

I wil be "off-topic" here for several days, but I promise you that if I find the link again, or anything similar, I will post it here right away

Nice thread, BTW, I'm mostly just following, you are "The Boss" here. > barryhc

[ChemE]

It seems that most DIY'ers here at RC are familiar with the fact that some proteins need up to 120 seconds to become trapped at the bubble wall but I think the everyone is applying it incorrectly. We do calculations for flowrates in order to ensure that water spends at least 120 seconds in the skimmer but I think that what we really want is for bubbles to spend 120 seconds in the skimmer.

Think about it, the protein is in a differential volume of water which is moving very slowly down the reaction chamber. All the while it is being bombarded with very very fast moving bubbles. This doesn't result in 120 seconds of contact time with a bubble wall.

The alternate scenario is the same protein in the same differential water volume travels to the bottom of the skimmer and comes in contact with a bubble just as it is formed. Chemical engineers use the "no slip" boundary condition which theory predicts and experiments validiate. This no slip boundary condition states that flow at a surface is stagnant. Therefore the water right at the bubble wall remains at the bubble wall as long as that bubble exists. Once the bubble pops the water at its wall is released back into the bulk fluid and is mobile again. Don't believe the no slip boundary condition? Put talculm powder on your hood before you leave for work Monday morning and drive 1000 mph to work; it will still be there when you arive very very early. Anyway, if the bubble takes 120 seconds to travel up the skimmer the protein's weakly hydrophobic section has a chance to move into the bubble thus capturing it.

Now, I of course realize that 120 second bubble rise time calls for a 100' tall skimmer but perhaps a very very low turbulence recirculation design in which 90% of the bubbles gently recirculate in a wide diameter loop (minimal turbulence) is the ticket. The only real issue I see here is keeping a high enough downward flowrate in the recirculation portion that the bubbles are entrained by the water and carried down and around rather than just building up or running back up towards the water surface. If it takes a bubble 10 seconds to make the recirculation circuit and there is a 90% chance of taking the circuit then 0.9^12 bubbles will be in intimate contact with the water at their surface for 120 seconds. See where I'm going with this?

OK, since this is contrary to conventional wisdom let the dissent begin.

[tinygiants]

Interesting points there ChemE. I had not considered the info in that way. If the 120 seconds is a valid number, and the bombardment rate is a valid ratio, the water dwell would have to be 20 minutes.

Water dwell / air dwell = bombardment ratio (goal of 10 according to Aquatic Systems Engineering by Peter Escobal)

water dwell / 120 seconds = 10
Water dwell = 1200 seconds = 20 minutes

With a 2 gallon feed rate my skimmer would have to be 40 gallons big to have that water dwell. In a 12 inch body, that is over 80 inches tall. This of course does not guarantee the air bubble dwell.

There are so many factors when trying to balance everything.

Dale

[tinygiants]

Originally posted by barryhc
I think you are more up on this than I am "tinygiants", but again this is the ratio of oxygen to water in the "dwell chamber".

I think that I may have actually found this on a link within the links that you have provided here, I am not sure.

I wil be "off-topic" here for several days, but I promise you that if I find the link again, or anything similar, I will post it here right away

Nice thread, BTW, I'm mostly just following, you are "The Boss" here. > barryhc

"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."

I found the above quote at this website. Looks like you were right. My original info was wrong.