Development of a Method for Quantitative Evaluation of Protein Skimmer Performance

my head hurts... lots of reading, but the conclusion is that they all (at least the 4 tested) perform the same
 
Reading into that a little, it seems to point to the new (and pricey) recirculating designs if anything being less effective since they minimize the Q (flow rate into the system) and will tend to waste air and bubbles on already "depleted" water.

Their results make a lot of sense. The skimmer volume relative to tank volume, and the input flow rate are the key drivers to skimmer effectiveness. Bigger tanks need bigger skimmers, which allow you to have a higher flow rate without skimming too wet. If you want a small skimmer to be effective you need to crank up the flow rate which means wet skimming.

If the air injection method has little impact on effectiveness, that leaves concerns like power consumption, pump reliability, noise level, ease of cleaning, and the effect on micro fauna the key drivers for technology choice. I would love to see a test that tries to quantify if there is a difference between air driven and veturi skimmers verses the down drafts and needle/mesh wheels as far as chopping up the microfauna.
 
<a href=showthread.php?s=&postid=14176208#post14176208 target=_blank>Originally posted</a> by Piazzon12
my head hurts... lots of reading, but the conclusion is that they all (at least the 4 tested) perform the same
Click to expand...

The article can be a bit misleading in that regard. They did not really perform the same, except when they controlled for volume and flow rate.

The needle wheel unit sucked in straight performance, only getting 56% of the BSA compared to nearly 80% for the venturi and downdraft and 95% for the air driven.

The downdraft and venturi had similar absolute performance, but one was larger in volume and the other had a higher flow rate. Looking at the raw data also makes it fairly clear that volume of the skimmer plays a much stronger role than the flow rate. The downdraft needed about 65% more flow to make up for having 22% less volume. That is a bad trade off in noise and power consumption I would bet.
 
Agree with that m fine, very interesting results in all. Also seems to indicate that since performance is relatively level across the technology playing field, cost should be a big factor when choosing a skimmer
 
skimmers

skimmers

I use two different skimmers a needlewheel Asm3 and an air driven 6 foot homemade. So I get the best of both . RickM
 
some questions regarding the article

some questions regarding the article

Ken Feldman has graciously given me permission to post his reply to the questions that were raised in this thread regarding the published article "The Development of a Method for the Quantitative Evaluation of Protein Skimmer Performance" found here: http://www.advancedaquarist.com/2009/1/aafeature2



quote:
--------------------------------------------------------------------------------
Originally posted by bosborn1
Funny how they didn't include one "modern" skimmer in the test. Most of those use designs that have been trumped by todays skimmers. Why only use mid-grade skimmers? Why not throw in a R.E. Bubbleking or an ATB? Even perhaps a high end Beckett?
--------------------------------------------------------------------------------



Our skimmer choices were dictated by two concerns:

1) We only had a 30 gallon sample of reef tank water which came from a water change on my tank (total volume ~ 175 gallons). So, we felt constrained to use small skimmers rated for small water volumes. At the time that these skimmers were purchased (April 2006), we were unaware of any "high-end" skimmers that were appropriate for this small of a water volume.

2) Money. We operate on a very limited budget, and the cost of one "high-end" skimmer (large size) was more that all four of the small skimmers that we purchased.


quote:
--------------------------------------------------------------------------------
Originally posted by bosborn1
Most of those use designs that have been trumped by todays skimmers.
--------------------------------------------------------------------------------



What does "trumped" mean in this context? What criteria are you using to judge that one skimmer "trumps" another? Once you educate me on this point, perhaps I can comment further.


quote:
--------------------------------------------------------------------------------
Originally posted by bosborn1
Why only use mid-grade skimmers? Why not throw in a R.E. Bubbleking or an ATB? Even perhaps a high end Beckett?
--------------------------------------------------------------------------------



I'm not sure what "mid-grade" means. If mid-grade refers to price, see above. If by "mid-grade" your are referring to some aspect of skimmer performance, then you will have to explain to me what metric specifically you are using to measure "performance" before I can comment further. We plan to return to the question of skimmer performance with an RC80 from Euroreef. That skimmer uses an independent means to control air flow and water flow, so it will allow us to measure skimmer performance metrics at different flow rates. I note that the Bubble King mini is an appropriate size to test, but at almost $900, it may be a little too pricey for us.


quote:
--------------------------------------------------------------------------------
Originally posted by bosborn1
Also I noticed that their was no citing of P.R. Escobal. I always refer to Escobal's work when evaluating skimmer design.
--------------------------------------------------------------------------------



I am familiar with Escol's work. In fact, I have his book on my shelf. I did not reference it because Escobal's treatment of skimmers points to different aspects of skimmer operation that those that interested us. For example, our mathematical modeling was aimed at connecting a measureable figure-of-merit for skimmer performance (the rate constant k) to system parameters such as flow rate (Q), reservoir volume (Vr) and skimmer volume (Vs). As near as I can figure, Escobal actually ASSUMES a certain organic removal efficiency (the "9.2" in Eq. 9.1) that he calls a purity coefficient, and then proceeds to use that value to estimate other quantities that are of interest to him. These quantities primarily relate to the properties of bubbles in water ("bombardment rate", etc.). As an aside on this latter point, Escobal's treatment is hopelessly naïve. He does not take into account bubble coalescence, foam formation and drainage, etc., that all play a big role in determining overall skimmer performance. In addition, part of his physical model for how bubbles remove organic impurities from saltwater (see Fig. 9.2) â€"œ "like striking a dusty cushion with a stick" - is implausible, to say the least. If you are interested in mathematical modeling of the whole "bubbles-in-water" phenomenon, there are much more sophisticated and realistic approaches to be found in the references from the "History of Protein Skimming" section of the AA article.


quote:
--------------------------------------------------------------------------------
Originally posted by bosborn1
His book on skimmers is considered the bible on skimmer physics by most industry professionals
--------------------------------------------------------------------------------



?? I don't know how to respond to this claim. Perhaps I might just point out that Escobal proposes several models (= hypotheses) regarding skimmer performance (he calls them "laws", Chpt. 8), but he never performs any experiments or generates any data to test these hypotheses. In our AA article, we do generate data that tests our hypotheses (e.g., Figs. 9 and 12 = straight lines; Figs. 8 and 11; theoretical curve (green line) fits the actual data (purple squares)).
 
I'd like to add that I believe these are all valid questions and I expect more might be raised however I'd like to give Ken and company kudos for finally addressing the topic of how to quantitively evaluate skimmer performance.

For so many year reefkeeper would try to judge/compare skimmer efficiency on crazy things such the color of the skimmate, quantity of bubbles in the skimmer, height of the foam head, how quickly a foam head was formed, height of the skimmer, width of skimmer, cost of the skimmer etc. etc. etc.

This research was long overdue!
 
I'm glad to see that someone has published an article on skimmer performance. This is the first time I've ever seen anything on this and I'll admit that I didn't go thru the whole article, but who does. :D That's why there are abstracts.

They put a lot of work into obtaining this info and I wish they would've included more graphs comparing TOC/BSA and processing time. Only one (Euroreef experiment) was shown, but IMO this is valuable info. All the tables are great and full of good data, but the only thing I got out of the tables was that flow rate thru the skimmer probably doesn't affect the skimmer performance all that much.

I'm a little more interested in the TOC analysis and it would have been nice to see what the initial values were and how each of the skimmers performed in removing TOC over time. Would only 30% removal be obtainable because it flattens out or is it because the test was stopped at a set time and this was the value obtained as their endpoint?
 
<a href=showthread.php?s=&postid=14183826#post14183826 target=_blank>Originally posted</a> by chem-e
I'm a little more interested in the TOC analysis and it would have been nice to see what the initial values were and how each of the skimmers performed in removing TOC over time. Would only 30% removal be obtainable because it flattens out or is it because the test was stopped at a set time and this was the value obtained as their endpoint?
Click to expand...
Ken's reply:
#1. The remaining TOC is not removed by the skimmer. Some initial runs were carried out to 6 hrs just to make sure that nothing much was happening after the initial hour or so when the TOC was removed. 30% is really at the high end. A more realistic value of the % TOC removed is probably closer to 20 - 25%.

I will add parenthetically that GAC is far superior to bubbles in terms of TOC removal ability; we can strip reef tank water down to ~ 0.2-0.3 ppm TOC (around our detection limit) using GAC. Basically, my working hypothesis has to do with the type of chemical binding profile that GAC presents to TOC vs. that of bubbles. In other words, I suspect that bubbles are not really very good at binding/removing TOC; they are a lot cheaper than GAC, though, and that may be their primary advantage.

Also, as Randy Holmes-Farley has pointed out in another thread, the bubbles oxygenate and add CO2 to the water, and those benefits, while now bearing on TOC removal capabilities, are quite valuable in their own right.
 
This was long overdue. Great work. Now I don't feel so bad that I don't have a grand to drop on something that mixes air and water! Pick a reliable brand with good support and go with it. Thank you for this article. It is a great start to actually evaluating skimmers on things other than looks.
 
Yeah we might as well use a traditional air driven Sander skimmer. Come to think of it I would like to get one now, just to compare it with my other skimmers. I hope they do a skimmer/ozone test next. Do you guys need donations?
 
I really was surprized by two results:

a) How similar their performance is (At least at the manufacturer's recommended settings) regardless of the mechanics for bubble formation.
This could have a large impact in the decision making process as now things like power consumption, initial cost, easy of cleaning and maintenance, physical size vs capacity and simplicity have a major impact.

b) How fast the skimmers may run out of stuff to remove contrasted to the usual expectation of haveing them generating skimmate continuously. This may confirm that although some skimmers may be a bit faster the majority will skim out most skimmable matter in a short time 1 to 2 hours confirming that in reality it is not possible to overskim.
 
This thread should get moved to the Equipment forum. Sanjay had a link to this thread there and that is the only reason I would have ever seen this. There are many people out there "testing" different skimmers by looks, air draw, and pump wattage. I would bet they would send in, temporarily of course, some of thier skimmers to be tested in a more scientific fashion. I think it would be good to compare skimmers with the "latest technology" of mesh mods and bubble plates.
 
<a href=showthread.php?s=&postid=14205119#post14205119 target=_blank>Originally posted</a> by jdieck

This may confirm that although some skimmers may be a bit faster the majority will skim out most skimmable matter in a short time 1 to 2 hours confirming that in reality it is not possible to overskim. [/B]
Click to expand...

Actually, the rate of TOC addition to the tank's water as well as the removal rate must be taken into consideration to evaluate whether "overskimming" is possible. In addition to skimmer removal of TOC, both GAC and bacterial action (feeding) remove TOC as well, and those removal mechanisms should be factored in also when evaluating the net TOC budget. I have some pretty good ideas about GAC removal (see http://www.advancedaquarist.com/2008/1/aafeature1/; http://www.advancedaquarist.com/2008/2/aafeature1/) and bacterial removal (see http://www.advancedaquarist.com/2008/8/aafeature3/, Fig. 6), but I don't at present have a good way for measuring the rate of TOC addition from various sources (feeding, phytoplankton and coral metabolism, etc.)

Ken
 
Very useful and informative study. Thankyou.

TOC removal( at 25 to 30%) is a concern particularly in view of the generally held assumptions about skimmers removing it when carbon( vodka, etc) dosing is in use.
 
Wouldn't it be something that you could at leat get a bead on by using a FO system, bare bottom and measuring food input? I mean fish can only poop what they eat, and that eliminates light "eating" creatures by and large the spare diatom, and alga aside...

Tracking how much food you add, and what your TOC levels are over time should give you a decent graph if the variables are at least minimal... And with that you could figure out the value of over skimming if any.
 
I'm not sure about the strength of the connection between food in and TOC out for a reef tank. On this point, I'll offer one odd observation that surprised me, and that I don't really understand (not published in the AA article):

I feed my tank a very regular amount of food on a very regular basis on purpose. In fact, I have dried my food portion under vacuum just to see how much it weighs, and find that I feed my tank about 4.2 gms of solid food per feeding, 3x per week.

If I carefully collect my skimmate over the course of a week (3 feedings = ~ 12.6 gms in total) and dry it out under vacuum, I get 25 gms (3.6 gms/day).

If I DO NOT feed my tank and collect the skimmate, dry it and weigh it, I get 17 gm over 4 days (4.2 gms/day).

I skimmed as dry as possible for these experiments, so I collected less than ~ 50 mL of water (with its salt component) in these skimmate sludges.

I've only done these experiments once, so there are no statistics, but I suspect that the weight of skimmate produced per day would converge to be about the same with or without feeding.


So, the implication is that food addition to the tank doesn't significantly influence the amount of skimmate produced - corals produce a tremendous amount of TOC that feeds the tank's microbes, and perhaps all I am really skimming out is some part of that part of the food web (corals <=>TOC <=> microbes) not involved with fish, fish poop, etc. I really don't know. I'll have to reproduce these experiments to get good statistics at some point (collecting, drying and weighing skimmate is really disgusting, so I hope to get my research assistant Kelly to do this over the summer!).

Ken
 
You must log in or register to reply here.

Similar threads

Tamara Marshall
Comprehensive Article About Protein Skimmers- Manta Systems Article
Replies
0
Views
254
Tamara Marshall
Tamara Marshall
C
New Protein Skimmer
Replies
6
Views
863
griss
griss
B
Chemiclean & protein skimmer
Replies
9
Views
954
bcwalz
B
L
4 outdoor Salt Aquariums Recirculation Treatment Design Question
Replies
0
Views
307
lvega2020
L
M
ATO Problem in JBJ NanoCube 24g
Replies
0
Views
236
MNtadpole
M
Back
Top