Scaling Up BSF Production | Theoretical and Practical Effect of BSF Bin Space Surface Area and Food Scrap Load Rate on Larval Yield : The Life and Times of BSF (Black Soldier Flies)
DipTerra LLC      
HomeAbout UsProducts & ServicesPublicationsForumsQ&A'sBlog IndexBlog

image of DipTerra logo
Insect-based Processing of Biodegradable Waste
Contact Us

Scaling Up BSF Production | Theoretical and Practical Effect of BSF Bin Space Surface Area and Food Scrap Load Rate on Larval Yield

by Terry Green on 06/16/15

Whereas Black Soldier Fly (BSF) larvae can be easily grown and harvested on a small scale from food scrap and animal manure (see Black Soldier Fly Processing of Biodegradable Wastes, and The Black Soldier Fly, Hermetia Ilucens, as a Manure Management / Resource Recovery Tool), scaling up production is far more challenging. This blog reviews some of the characteristics associated with BSF larval behavior, current theoretical limits on larval yield inherent in growing larvae in processing bins, and issues relating to waste loading rates bearing on larval yields in scaling up production of BSF.

Scaled up production requires steady-state input and processing of large quantities of waste, establishment of a stable and sustained method of propagating larvae on site where larval processing occurs, harvesting and marketing larvae raised, and safe handling and disposal of all waste byproducts generated, year in and year out. These operations must be carried out economically and efficiently while generating an income stream covering the cost of the operation. In this regard, it is important to understand the relationship between waste surface area in farming larvae, the average daily waste load rate (WLR), larval behavior and yield in planning  and fact checking  the metrics and layout of a BSF farming operation.

Consider first how BSF larval behavior and average WLRs interplay in influencing larval yield.  At DipTerra we have observed over the last four years that yields to a good approximation are directly dependent upon the surface area of the waste.  This likely reflects larval feeding preferences in growing off of waste. Larvae, for example, respire air through spiracles projecting outward from the segments of their body as they feed on waste. Because they are phototropic (light fearing), yet also need access to oxygen as they grow off of the waste, they tend to aggregate in waste just under and nearest to the waste-air interface, rarely burrowing deeper than about four to six inches (~ 10 to 15 cm) beneath the surface of waste.

The consequence of this feeding pattern is that larval production, and hence larval yield, to a very close approximation depends principally on larval bin surface area. The total bin surface area of a farming operation, therefore, in combination with the WLR, are two of the most important determinants (aside from proper maintenance of environmental conditions in farming larvae in bins) in scaling up larval production.

Identifying and managing a proper WLR in feeding and propagating larvae requires careful observation of the consequences of loading waste into processing bins. Loading a processing bin with waste at a rate exceeding the capacity of larvae to fully assimilate it, for example, creates “dead spaces” where unprocessed waste piles up. Under these circumstances larvae die off trapped in waste piled too deep beneath the waste-air interface. Carbon dioxide accumulation, combined with excess organic acid production, formed in these “dead zones” (actually anaerobic pockets of waste, see Recycling Food Scrap? | What’s with the Stink?) furthermore can drive even young larvae prematurely from processing bins. Choosing to work at too low of a WLR relative to the larval rearing capacity of a bin, on the other hand, starves off larval poduction.

It is possible given familiarity with the properties of the waste used in farming BSF larvae to calculate in advance the maximum theoretical yield of larvae which could be harvested per unit time interval at any waste loading rate (WLR, the wet weight in Kg waste per day fed to larvae).  Two additional variables required in calculating the theoretical larval yield at any WLR selected are: (i) the average moisture content (MC, expressed as wet weight minus dry weight divided by wet weight) of the waste, and (ii) the average larval feed conversion ratio (FCR, the dry weight of harvested larvae divided by the dry weight of the waste).

If these variables are known, one can calculate the theoretical upper limit in larval yield for any WLR used in scaling up a farming operation. The equation is as follows:

Maximum Theoretical BSF Larval Yield (dry wt. Kg/m2/day)

= (FCR)(1 – MC)(WLR)

The FCR, itself, will vary depending upon the source of waste, its nutrient content and digestibility, how well the larvae are able to feed off of microbes and byproducts produced as the waste decays, and by many other complex variables. It must by necessity be determined empirically.

Because the bio-conversion of waste into insect biomass, itself, consumes chemical energy in building and assembling critical cellular structures, protein, exoskeleton, etc., and because energy is also consumed as larvae crawl about feeding on the waste before self-harvesting free of the waste, the FCR expressed as the ratio of larval dry weight harvested to dry weight of waste needed in producing the harvested larvae is always less than 1.0.

The FCR for BSF larvae grown off food scrap, calculated from data generated at our DipTerra plant (see BSF Metrics & Yields| Scale Up Production of Black Soldier Flies), and from two other independent groups studying BSF larval conversion rates (see Conversion of organic material by black soldier fly larvae: establishing optimal feeding rates and Bio-conversion of Putrescent Waste) all indicate that the conversion ratio in growing BSF off of food scrap is ~0.25.

Concerning an appropriate WLR, experiments by Diener et al. (see Conversion of organic material by black soldier fly larvae: establishing optimal feeding rates) suggested that the optimal WLR (which they call “the potential daily feeding capacity”) is in the range of about 3 and 5 Kg/m2/day which closely matches a WLR of ~5 Kg/m2/day derived in our own independent studies in growing BSF larvae off of food scrap waste  (see BSF Metrics & Yields| Scale Up Production of Black Soldier Flies). ESR in yet another study has suggested that the WLR is closer to 15 kg/m2/day (see Bio-conversion of Putrescent Waste). 

It is not clear if over the long haul the higher WLR suggested by ESR using food scrap waste would prove sustainable. Our observations with food scrap indicate that this is too high of loading rate to sustain without encountering a long term decline in the average larval yield.
graphical image of BSF larval yield versus food scrap loading rate
Fig. 1. The predicted effect of food scrap (waste) load rate (WLR, kg/m2/day) on the theoretical maximum possible harvested yield of BSF larvae (dry weight, kg/m2/day). 
*Calculations are based on an average food scrap moisture content (MC) of 80% water and BSF larval feed conversion rate (FCR) of 0.25. (Copyright © Terry Green 2015, all rights reserved).


By using a value for the FCR of 0.25 in the formula above, and considering that the average moisture content of food scrap is approximately 80% (see, for example, Food Waste Directory - Maceration & Dewatering), one can calculate the maximum dry weight BSF yields attainable in Kg/m2/day as a function of WLR (see Fig. 1).  Since the average moisture content of prepupae harvested from food scrap is ~0.5 (see Commercial Production of Black Soldier Flies |Preserving Harvested Larvae), the maximum yield in wet weight units of BSF larvae harvested off of food scrap can also be calculated to be roughly twice that displayed at the varying WLRs shown in Fig. 1

Even though these theoretical calculations expressed in Fig. 1 appear to suggest that one could achieve higher larval yields at higher WLRs, it is important to keep in mind that these calculations represent the highest yield possible at a given WLR assuming that the loading rate does not exceed the capacity of larvae in turning over the waste as rapidly as it is delivered. However, as noted earlier, overloading a bin with waste leads to a buildup of unprocessed waste that inevitably causes a decline in larval yield.

Once one has selected a WLR and feedstock waste to be used in farming BSF, an approximate scaling of an operating plant can be roughly mapped out for a targeted annual production level of BSF larvae sought. For example, suppose you want to calculate the approximate size of a farming operation needed in producing 100 metric tons of BSF larvae (dry weight) annually.  If you chose to load food scrap into larval bins at a WLR of 2.5 Kg/m2/day, the total bin area needed in harvesting 100 metric tons of BSF larvae per year would be ~ 100/365 x 1000 divided by 0.13 (the calculated upper theoretical larval yield possible per square meter of waste), or  ~ 2222 m2. The WLR in food scrap waste required in meeting this production level would be 100/0.25 x 5, or 2000 metric tons per year (~ 5.5 metric tons per day). For a WLR of 15 Kg/m2/day, the total bin space needed would be ~ 317 m2.

While working at a higher WLR clearly could reduce the footprint of the farming operation, the issue is whether a farming operation would be able to manage larval bins and yields year in and out at the higher loading rate. So far based upon our own experience, WLRs in excess of 5 Kg/m2/day have not proven sustainable. While leading initially to higher larval yields, over the long haul the higher loading rates have consistently proven to negatively impact long term larval yields.

Check back for more to follow on the management and strategies in farming BSF. Comments on this blog, or any of our other blogs, are always welcome. Follow us through our RSS feed. For additional information or follow-up questions, visit our Q&A's or Forums page, or Contact Us (http://www.dipterra.com/).


Comments (6)

1. daniel michaelson said on 7/6/15 - 10:24PM
Thank you for your analysis. So at your optimal feeding rate of 5 kg/m2/day you would need a total surface area of 11,790 ft2 to get 100 tons of larva worth about $200,000 as fishmeal or an order of magnitude more as pet food. Seems reasonable enough. As a side note, I have been using the BSF as treatment method for my humanure. They eat it as fast as I produce it and I feed the larvae to chickens without any problems. Thanks for the great work!
2. Terry Green said on 7/8/15 - 10:29AM
Daniel, I agree with your summation. Keep in mind that there are a couple of other variables which may be exploited to increase larval yield linked with increasing the turnover rate of waste fed to the larvae. The calculations described in this blog project larval yield per unit processing area based upon an average FCR of 0.25 (expressed on a dry weight basis) and a WLR in units per day fed to the larvae. In principle the WLR also depends to some extent upon how the waste is preprocessed before delivery. It may also be increased to some extent by amplifying the larval population per unit processing area. Stay tuned - I plan to follow up in coming blogs on some of these latter issues concerning scaling up BSF production.
3. Ilkka Taponen said on 7/16/15 - 03:11AM
Hello! Thanks for the very interesting blog! How does the quality of the organic waste affect the larvae growth? E.g is there a significant difference if the waste is high in protein or something that is very wet and poor in protein? ps. Check out my blog on the topic from ilkkataponen.com
4. Terry Green said on 7/17/15 - 05:25PM
Thank you,Ilkka. The nutritive value of waste (protein, and additionally the type of carbohydrate, polysaccharide, mineral and fat content, etc.) all absolutely affect insect growth and viability. Larvae grow better on food scrap, for example, than manure. From a practical perspective in scaling up larval production, it is most important to deliver the waste in a fine pulp slurry in maximizing access to its intrinsic nutritive value. I will be posting more on this subject in coming blogs. Thank you, too, for sharing your blog. I agree with much that you have written. Concerning issues relating to commercial insect farming and the prospect of cultivating potential pathogens, there are a number of well-established ways of safely managing scaled up insect farming in mitigating this potential risk. The important point which you also noted is to be aware of the potential risk, and to put into practice measures for monitoring and mitigating these risks so that it does not become a problem.
5. Irwan Santoso said on 4/26/18 - 03:05AM
Hi, I may have missed it, but for the ideal WLR of 5 kg / m2, what is the ideal larvae population per m2 or per sq ft?
6. Terry Green said on 4/27/18 - 10:24AM
Irwan, it is better not to impose an artificial limit on the WLR or larval density. There is no obvious upper limit on the density of larvae that can be grown in a bioreactor except at the physical limit of larvae packed side by side. The optimum WLR is a relative term and is largely dependent upon the design of a bioreactor, its layout and practical considerations in the management of waste fed to larvae. A reliable (“optimum”) WLR is about 5 Kg/m2. This WLR can be substantially increased with experience and better management techniques. Larvae adjust to the conditions and availability of waste fed to them. With proper waste aeration, management of the waste temperature and design elements that allow larvae to move freely in and out of a bioreactor, there are no adverse outcomes on the growth of larvae at higher WLRs. Instances where larvae fail to thrive are largely associated with mismanagement of the waste and ensuing stressful conditions setting up which limit their growth.


Leave a comment


Dipterra's Blog - The Life and Times of BSF (Black Soldier Flies)