Rhodimet Itw techno Homogeneity - Marc Perel, Solution Application Manager, at Adisseo comments about homogeneity and recovery of additives and especially liquid methionine in feed.
Marc, In my opinion a CV = 10% is no good enough. It means that the additive is present in the feed at plus or minus 20% of the average intended dose. I don't think this is acceptable, at least for poultry. A CV = 5% means that the distribution of the product is between plus or minus 10% of the average; this is what I consider a satisfactory distribution for an additive.
Dear Ricardo Hume Thank you for your comment and your interest in this video. In order to avoid any misunderstanding regarding the interpretation of the CV, let me clarify that, considering a normal distribution of the additive in a batch of feed (Gauss curve), a coefficient of variation (standard deviation/mean value) of 10% means that 68% of the mass of the batch contains between 90 and 110% of the mean value, 95% between 80 and 120% and 99% of the mass of the batch contains between 70 and 130%. A value of 10% is commonly accepted in animal nutrition. For example, the European Feed Manufacturers Guide established by the FEFAC mentions that “a target maximum percent coefficient of variation (CV) and mean percent recovery must be established taking into account the analyte, the target levels and background values. In most cases, a target CV of less than 10% should be achieved.” I am not a zootechnician, but I would say that the criticality of the heterogeneity of an additive depends on several factors: notably the nature of the additive and its rate of incorporation in relation to the needs of the concerned animal (and the daily feed intake). An elephant may be more tolerant than a shrimp... In any case, and I agree with you on this point, it is up to each manufacturer to determine its own quality standards, which may go beyond the good practices codes and regulatory requirements. Kind regards, MP
Marc, see what other experts think about a CV = 10%.
All About Feed Magazine
Wayne Cooper
Feed mill and machinery expert
Expert opinion
18 May 2020
The CV of
Ok, so this is a pet peeve of mine. This standard has been in place since 1985….forever! Is it not time to step up our game? I had actually thought our major challenge would always be medications. However, as medications are cut back, we are now getting multiple products to replace them. Enzymes, probiotics, herbs, you name it. Whatever it is, it is concentrated to small addition rates. We need to evenly spread it throughout the batch. At one of the mills I supervised, we were operating a “5 ton mixer” at 110 tph and CV’s under 4. If you are looking
at a new mixer for any reason, do you want to go with “OK” CV’s of <10 for the next 30 years? you can do better if you check around.
Dear Ricardo Hume
I do not want to give the impression that there is any disagreement between us or even that I am downplaying the importance of the homogeneous distribution of additives in food, quite the contrary! My speech is as follows: monitor the performance of the mixers regularly (every year), set limits for the recovery rate and CV, implement corrective actions if the results are outside these limits. A simple quality approach in short.
My job is precisely to bring nutrition and technology closer together and your interest in what happens in the plant is very important in this respect.
Concerning this CV limit, by default I say 10% (common reference in regulations and quality standards), you say 5%, it's fine. And obviously, more demands will be placed on a new installation than on an installation that has been in operation for several years.
In any case, my colleagues and I are always available to help our customers achieve their quality objectives, in any context.
Yours sincerely,
Interesting comment about the CV's of mixer tests and the 1985 date for the 10 % CV acceptability. While employed at Degussa, now Evonik in the 1980's mixer profliles were offered as a customer service. At that time a standard was needed to compare for % CV's for mixer profiles. The standard at that time, around 1985, was the Kansas State research on mixer profiles using the QuanTab salt assay. This data indicated that a good feed mixer in the 1980's could achieve a CV of 10% using this somewhat variable assay. Mr. Perel's comments are very well stated and very good for statistics and establishing goals for probably of having a certain nutrient level..
However, as Mr. Hume points out we can and should do better than the old standard of 10%. A new standard from my experience with assaying for added methionine in a good laboratory for assays is a CV of 2-3 % above the variation of the assay being conducted.
In my experience the added methionine assay has an assay variation of 2-3 % for the assay itself ,assaying 20-30 samples over a period of several days. Thus, a very good feed mixer with an addition rate of 4-6 pounds of methionine per ton should have a CV of 5-6%. Not the old 10 % CV that was established with the variable assay in the late 1970's- 1980's.
The feed mixers that I used achieve this performance level for methionine additions. The methionine assay is a good assay for the general condition of the mixer because the added test substance is added in the range of most micro feed additives 1-8 pounds per ton and thus test the mixing of ingredients added a lower addition rates and the bulk density of methionine is at a mid point of most feed ingredients other than minerals.
Minerals with their higher density should be tested with a different assays . Calcium is a difficult nutrient to achieve good CV's for in diets. Part of the problem is that we tend to use the same laboratory tests to assay for calcium in ingredients that have levels of 0.1% calcium or less to finished feeds that have levels of 0.8 to 3.5 % calcium and compound the assay problem by using either finely ground limestone or coarse limestone particles as the limestone source in these feeds. The solution is the same as above, optimize the assay for the amount and form of nutrient source (calcium) being added and the addition level added and then conduct the mixer profiles.
Dear Sirs, thanks for info. In case of liquid products spray methods, CV=10% is ok. But when we mix powder to powder, we have much more problems. At first, this is ditterent density weight of additives. At second, this is transportation. Vibration of cargo often caused mix stratification. Often is necessary to make re-mixing after transportation, before to feed animals.
Dear Fiodor. If I want to add 4kg of liquid methionine / ton of feed with a CV=10%, that means I am willing to accept a dosage range of 3.200 - 4.800 kg (considering plus or minus 2 standard deviations). This is a very broad range in my opinion. I still think that the CV must be 5% for liquid products as well.
Please let me comment your last message to avoid any confusion:
On one hand, we speak about homogeneity of repartition of an additive in a batch of feed, where the CV translates the variability between N samples taken from a same batch. I mentioned that a CV of 10% is acceptable, you prefer to say 5%, it is OK, I will not comment further. On the other hand, you mention "...a dosage range of 3.200 - 4.800 kg", addressing the dosage accuracy, which is another topic. Usual technical means (by scaling or flowmetering) generally allow accuracies of 1% or less to be achieved. Both notions (homogeneity & dosage accuracy) are complementary to reach the conformity of the finished feed. Best regards, Marc
Marc Perel Thank you Marc for your comments. I totally agree with you that a dosing accuracy of 1% is excellent. This means that when you add 4kg of a product / ton you will find samples between 3.960 kg and 4.040 kg/ton. That range represents a CV = 0.5%. Best regards, Ricardo
Ricardo Hume Dear Ricardo, 100% agree with your opinion. CV 5% is easy to reach, because if we will spray methionine into the mixer, during the mixing process, feed mix becomes to be (theoretically, if mixer is well adjusted) 100% homogenic, in splite of different size and specific density of ingredients. After the end of mixing process, during feed transportation, stratification is possible, but it is any case not stratification of methionine, because different feed ingredients, after spraying, already have the same methionine content! In case of pelleted feed, stratification is no problem at all.
In response to Dr. Marchenkov. Sorry, but I disagree very strongly! Animal feed mixing is very predominantly a dry mixing process. As soon as you add small quantities of ANY liquid into this dry mix you start to get "balling" where dry particles aglomerate. This immediately reduces homogeneity. As well as the basic physical mixing problems of adding liquids to a solid mix, we have the additional problem of the application. Spray of small liquid particles is best; but the ability to do this is again confounded by the dust within the mixer environment. Obviously some liquids MUST be added; fats, maybe enzymes, molasses in ruminant feeds. But addition of liquids should certainly be kept to a minimum. MPS
Martin Smith Does this imply that the post application technologies of especially feed enzymes has not reaped the anticipated results from a broader perspective? Many companies, for example BASF, have opted to produce more heat stable products than to rely on PPA technologies.
Hi, Paul. As you say, post pelleting of enzymes, esp Phytase, is carried out to avoid possible denaturation of the enzyme protein due to high temperatures in the conditioning and pelleting processes. Therefore this is not strictly mixing; and as such its homogeneity should be assessed independently of mixer profiling. Addition of heat stable enzymes to the mixer carries the same risk as I stated previously if they are liquids; however, so far as I know, many heat stable products are solid, so the risks associated with liquid application do not apply. MPS
Certainly an interesting discussion, I appreciate everyone's insight. As for the debate around the % CV, I think one of our issues is thinking in absolutes.
As Marc mentioned, the "standard" continues to be 10%, which is not to say that facilities shouldn't strive for even better homogeneity as technology improves. However, due to the variable precision of some assays (as Dr. Wicker describes) it is difficult to suggest that all %CV should be at lower levels. For example, assays for medications typically have high variability, and as they are of utmost concern to regulatory officials, demanding even lower %CV values may not be practical for the industry. On the other hand, for more precise assays, a facility may set a lower value as it's target.
The discussion around homogeneity and dosage accuracy is also interesting. Again, as Marc mentioned, they are complementary concepts. But it is certainly possible to have either the correct amount of an ingredient added and have poor mixing, or to have good mixing but the wrong addition level. I always encourage facilities to closely examine both the %CV and the average values; a 4% CV with a >10% deviation from the expected average recovery may tell us that mixing is fine, but batching/weighing is not.
I also advise facilities to test %CV on markers coming from various addition locations. If a facility uses some combination of hand-adds, micro system, tote system, and liquids, periodic assays should be run on something being added at each location to evaluate things like dead spots in the mixer and/or poor spray patterns.
Finally, and again as Marc mentioned, the species and production class is worth considering. If we are manufacturing a chick starter, evidence and common sense tells us that the feed should be very homogeneous given the meal size. But if we are feeding finishing hogs, less uniformity may be acceptable. Perhaps this would allow a facility to reduce mixing times for some feeds, and therefore increase them for others where uniformity is critical.
Ricardo Hume Not specifically tied to validated CV's, but yes, I have run across an instance where a facility has strategically reduced mixing times for certain formulas to create flexibility for increasing time on others. I don't necessarily like it without having the CV values, but if they could be validated and shown not to have a negative impact on animal performance, it could be a feasible data-driven strategy. Regardless, it would be best, and perhaps necessary from a regulatory perspective, to stay within the 10% envelope. Overall, I'll always recommend attaining the lowest %CV possible given available equipment and production requirements, but in some facilities with smaller and/or older equipment, such strategies may be a possible alternative.
Adam Fahrenholz Dr. Fahrenholz, May I disagree with you on less uniformity ifor feeds for older animals that have a larger meal intake as compared to a chick with a smaller meal intake. I believe that we are not only meeting the animal’s needs, but also we are feeding and altering the entire biome in the animal. A larger animal has the body mass to tolerate variations in meal intake as compared to a small chick, however the animal and it’s microbiome are rapidly affected regardless of size. It has been my experience that very consistent nutritional assays in the feed < 10 %, combined with quality ingredients, consistent feed particle sizes, and consistent feed manufacture results in healthy, highly productive animals with a stable productive biome whether the animal has a small or large meal intake.
David Wicker Ph. D. I agree with remaining below 10% CV, whether it is an arbitrary value or not, as the historical data and precedence supports it. I am referring more to a scenario of achieving 4% rather than 8% with an additional 20 seconds of mix time and whether those 20 seconds may be leveraged in some cases. As you said, larger animals are likely to tolerate this and I believe some (relatively old) data supports that conclusion. I'm not certain if there is any such data looking at the mocrobiome, but it would be interesting to evaluate. I suspect the degree of effect would be dependent on ingredients involved (e.g. the inclusion or non-inclusion of medications and/or probiotics) and their inherent mixibility dependent on product characteristics.
Dear Ricardo,
Very interesting point of view, I am completely agree with you that we should improve the homogenity of additives in feed. I think we should go further than a CV of 10%.
I have never met a person with your experience in animal nutrition, feed formulation and feed mills, and the most important is that you have always shared your knowlegment, this is very valuable and not easy to find in this market. You have been working in animal nutrition for so many years and you have developed an extraordinary advantage from others, the voice of the experience.
I remember that you once told me that the mixing time is the heart of a feed mill, it sets the pace for all the other steps, starting with grinding. Plants where weigh, grind and mix, mills gets harmed every time it starts and stops, waiting for the mixing time.
As you said, I have never seen a feedmill plant changing mixing times between batches to allow for diferent CV's.
Kind Regards,
Agustín De Cristofaro
Everyone has covered this subject so well, I just have one comment. The mixer that we see most commonly in a feed mill is a horizontal ribbon mixer, or a paddle mixer. In both cases you need to be careful that the batch size is in keeping with the manufacturers recommendations. I have seen a number of cases where over filling of the mixer and under filling of the mixer caused a change in the time that it takes to get a good CV. If it is necessary to run batch sizes that are less then half of the mixer volume capacity, then you will want to run tests to determine the optimal time for the best CV. I have seen this most often in situations where there is variation in the bulk density of the recipes that are being formulated. A ton of mids will take up more space then a ton of calcium carbonate. Try to pay more attention to the level in the mixer rather then the weight going into the mixer. Usually the optimal mix and mix time is based on the swept volume of the mixer.
Possible Issues of adding liquid to the mixer is not just CV%.
- longer dosing time: as we cannot shot liquids to the mixer like something we do for dry products.
Shorter application time of liquids typically produces a larger droplet size, which may lead to greater clumping tendencies in the feed and less uniformity of liquid incorporation.( please see : The Effect of Liquid Application Times, and Mixer Types with Different Wet Mix Times on Uniformity of Mix https://newprairiepress.org/kaesrr/vol2/iss8/42)
- longer mixing time as two process is involved 1) absorption liquid by dry particles and formation of bigger particles (it is function of viscosity of the liquid and temp.) , 2) Mixing of absorbed liquid : just started after abortion of the liquid by dry particles.
When dry particles come in contact with a large droplet size, they have a greater propensity to clump, which may reduce the uniformity of the total mix. While some clumps were reduced by the shear force generated by the turning shaft of the mixer, this is dependent upon other factors, such as mixer type and wet mix time. (Please see: Tekchandaney, J. K. 2013. Selection of solid blending equipment. Powder and Bulk Solids Exhibition & Conference, p. 1-5. )
- But it is not just issue with dosing and mixing time. You can manage it if you have proper mixing and dosing equipment and enough time. One main issue with liquid addition to the mixer is recovery rate. I did couple of mixer profile tests with liquid products. CV%’s were in range of 2.18% to 27.9%. This means if you consider right mixing and dosing equipment and procedure, you will get optimum CV%. But in all tests I did, the recovery rate were not in acceptable range (they were in rage of 76% to 86% while acceptable range are between 90 to 110% ) .
With application of liquids in the mixer, Lump formation on the mixing equipment like paddles or ribbons and … is almost unavoidable. Also controlling the quality of spraying (angle and size of the droplets of the liquid) is challenging because very soon dust materials stick to the nozzles.
I also 100% agree with Dr. Marchenkov’s concerns with segregation after mixing and/or different mixing properties of the dry materials with different specifications like particle size, density, dustiness and so on. That’s why quality of a dry additives also need to be taken under consideration (please see: Methods to evaluate handling properties of feed and feed additives , Feed Tech. 9.10 2005 – www.agriworld.nl , p 23-26 )
This published content is intended for Adisseo customers and prospects. Our policy is not to debate on forums or social media. Our teams are available worldwide to discuss with prospects or customers on methionine topics to deepen subjects published on various platforms, so do not hesitate to contact us via our website: www.adisseo.com.
.jpg&w=3840&q=75)
.jpg&w=3840&q=75)