Extraction rates and loss

[gclawson1]

Been digging into the data that was used for the extraction rates and crop loss. The methods from the food footprint paper say this (SI methods page 64):

We used extraction rates from the FAO commodity trees in the Technical Conversion Factors’ document131. These trees represent all processes a crop can undergo, and the percent extraction/yield at each step. Other sources were used to determine extraction rates for soya and soya protein concentrate132, pea protein concentrate133, and FOFM130.

For all fed animals, we calculated for each country, c, the total tonnes of each crop and forage fish they consumed given the heads of animals (livestock) or tonnes of production (mariculture, weight at slaughter), Ssw; the consumption, C, based on consumption rate for livestock (tonnes head-1 year-1) and feed conversion rates for mariculture (tonnes feed/tonnes product); and feed composition, adjusted for loss during processing, FCadj: Fc,f = Sswc X Cc X FCadjc,f

And the data they used for extraction rates and crop loss looks like this:

I'm having trouble figuring out how they calculated the crop_loss column. I assumed the crop_loss would be 1 - ext_rate, but that doesn't appear to be the case. They don't seem to have that information anywhere, unless I'm missing something. In the scripts, they use this method to figure out total tonnes of a product consumed in a country :

  mutate(tonnes_product = ingredient_demand * 100/(100 - crop_loss)) 

Just flagging this to remind myself to ask Mel about this when she returns from leave next week, cause I'm stumped by it.

[cottrellr]

Interesting OK.

I agree the difference between the extraction rate and the crop loss has me stumped too. But I assume that information is embedded somewhere in the processing variable? Would be interested to hear how the crop loss was generated once you find out. The 100/(100-crop_loss) bit is the raw material conversion factor so would be good to know how it compares to conversion factors.

Btw, I stupidly had soybean meal, soy protein isolate, and soy protein concentrate down as coproducts in my conversion factors calculation but they are parallel products that can be produced instead of each other from the same root material. So I have had to adjust. It is the soy oil and lecithin that are down now as coproducts for SBM and SPC and this raises the conversion factor. So you will need to refresh these if you bhave used them in other calculations. That’s limited to SPC and SBM though.

---

Richard S. Cottrell Research Fellow in Aquaculture Sustainability Institute for Marine and Antarctic Studies College of Sciences and Engineering University of Tasmania

Theme Co-Lead, Sustainable Futures and Planetary Health Centre for Marine Socioecology University of Tasmania

Size Ecology Labhttps://www.sizeecology.org/ | Centre for Marine Socioecologyhttps://marinesocioecology.org/themes/sustainable-futures-and-planetary-health/ Google Scholarhttps://scholar.google.com/citations?user=1pLCMKIAAAAJ&hl=en | ORCIDhttps://orcid.org/my-orcid?orcid=0000-0002-6499-7503 | @RichCottrell22https://twitter.com/RichCottrell22

On 29 Mar 2023, at 9:52 am, Gage Clawson @.**@.>> wrote:

Been digging into the data that was used for the extraction rates and crop loss. The methods from the food footprint paper say this (SI methods page 64):

We used extraction rates from the FAO commodity trees in the Technical Conversion Factors’ document131. These trees represent all processes a crop can undergo, and the percent extraction/yield at each step. Other sources were used to determine extraction rates for soya and soya protein concentrate132, pea protein concentrate133, and FOFM130.

For all fed animals, we calculated for each country, c, the total tonnes of each crop and forage fish they consumed given the heads of animals (livestock) or tonnes of production (mariculture, weight at slaughter), Ssw; the consumption, C, based on consumption rate for livestock (tonnes head-1 year-1) and feed conversion rates for mariculture (tonnes feed/tonnes product); and feed composition, adjusted for loss during processing, FCadj: Fc,f = Sswc X Cc X FCadjc,f

And the data they used for extraction rates and crop loss looks like this:

[image]

I'm having trouble figuring out how they calculated the crop_loss column. I assumed the crop_loss would be 1 - ext_rate, but that doesn't appear to be the case. They don't seem to have that information anywhere, unless I'm missing something. In the scripts, they use this method to figure out total tonnes of a product consumed in a country :

mutate(tonnes_product = ingredient_demand * 100/(100 - crop_loss))

I think I'll have to ask Mel about this when she returns from leave next week, cause I'm stumped by it.

— Reply to this email directly, view it on GitHubhttps://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/issues/4, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AJK3YJADIPPWSHI4AFRKRY3W6N2SXANCNFSM6AAAAAAWLHJU4U. You are receiving this because you are subscribed to this thread.Message ID: @.***>

This email is confidential, and is for the intended recipient only. Access, disclosure, copying, distribution, or reliance on any of it by anyone outside the intended recipient organisation is prohibited and may be a criminal offence. Please delete if obtained in error and email confirmation to the sender. The views expressed in this email are not necessarily the views of the University of Tasmania, unless clearly intended otherwise.

[gclawson1]

Alright I figured out mass allocation with the trade data, and have distributed the production for salmon feed on rasters! I'm missing soy oil and coconut oil from these, because I don't have yield rates for these yet, and I haven't accounted for fish trimmings yet.

Fish dominant diet (logged production):

Plant-dominant diet (logged production):

[gclawson1]

That was for crops... still need to work through the fofm!

[cottrellr]

This awesome, well done!

Richard S. Cottrell Research Fellow in Aquaculture Sustainability Institute for Marine and Antarctic Studies College of Sciences and Engineering University of Tasmania

Theme Co-Lead, Sustainable Futures and Planetary Health Centre for Marine Socioecology University of Tasmania

Size Ecology Labhttps://www.sizeecology.org/ | Centre for Marine Socioecologyhttps://marinesocioecology.org/themes/sustainable-futures-and-planetary-health/ Google Scholarhttps://scholar.google.com/citations?user=1pLCMKIAAAAJ&hl=en | ORCIDhttps://orcid.org/my-orcid?orcid=0000-0002-6499-7503 | @RichCottrell22https://twitter.com/RichCottrell22

On 31 Mar 2023, at 2:13 pm, Gage Clawson @.**@.>> wrote:

That was for crops... still need to work through the fofm!

— Reply to this email directly, view it on GitHubhttps://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/issues/4#issuecomment-1491261927, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AJK3YJGWFM3VF374IB4NTSDW6ZKWPANCNFSM6AAAAAAWLHJU4U. You are receiving this because you commented.Message ID: @.***>

This email is confidential, and is for the intended recipient only. Access, disclosure, copying, distribution, or reliance on any of it by anyone outside the intended recipient organisation is prohibited and may be a criminal offence. Please delete if obtained in error and email confirmation to the sender. The views expressed in this email are not necessarily the views of the University of Tasmania, unless clearly intended otherwise.

[bshalpern]

Mel is back online so definitely best to check with her. The crop stuff and accounting for feed was a huge pain in the ass and Mel spent a ton of time on it. She should be able to answer questions...

[gclawson1]

Updated scripts to include gross energy allocation. Also chatted with Mel and she thinks this all makes sense and looks good!

@cottrellr if we end up needing to account for processing stuff for FOFM, she mentioned it might be worthwhile to explore Jessica's trade stuff. They have intermediate (processing) countries included in their data, so we could probably work with that

[gclawson1]

Finally getting back into data prep!

We still have a couple of unresolved issues regarding yield conversion factors for allocation that I need. I'll be working this week to fix these.

1. Difference between pea starch and pea flour. The Aas et al. 2022 paper reports "pea flour", so that is the ingredient that I am using in my analysis. However, we currently only have yield data for pea starch.
2. Need to find yield info and coproduct yield info for coconut oil
3. Soy oil is only listed as a coproduct in the data that Rich uses. I'm using it as an ingredient in my analysis, so I need it to be an ingredient, and have co product yield factors (explained below).

Pea starch vs pea flour:

(I asked chatgpt for this description)

Pea starch is a fine, white powder that is extracted from peas through a process that involves washing, grinding, and separating the starch from the other components of the pea. Pea starch is a versatile ingredient that can be used in a wide range of food applications, including as a thickener, binder, and stabilizer. It is often used as a substitute for other starches, such as cornstarch or potato starch, in gluten-free and vegan recipes. Pea starch is also used in the production of processed foods, such as baked goods, snacks, and pasta.

Pea flour, on the other hand, is made by grinding whole peas into a fine powder. Unlike pea starch, pea flour contains not only the starch but also the protein, fiber, and other nutrients found in the pea. Pea flour is often used as a source of plant-based protein in a variety of food products, including bread, pasta, and snacks. It can also be used as a gluten-free flour substitute in baking, although its texture and flavor are different from wheat flour.

So pea starch is a material that is derived (or could be derived) from pea flour.

Coconut oil

We just don't have any information on this yet, so need to track it down! Another chatgpt generated explanation of co-products and coconut oil:

The production of coconut oil typically generates several co-products, which can have various uses and applications. The specific co-products can depend on the extraction method and processing techniques used. Here are some common co-products associated with coconut oil production:

Coconut Meal: This is the residue left over after the oil has been extracted from the coconut meat. It is a high-fiber, high-protein product that can be used in animal feed, particularly for ruminant animals like cows.

Coconut Milk: This is the liquid that is extracted from the coconut meat before oil extraction. It is used as an ingredient in many culinary dishes and as a base for many coconut-based products like coconut cream, coconut yogurt, and coconut ice cream.

Coconut Water: This is the clear liquid found inside young, green coconuts. It is a popular beverage, especially in tropical regions, due to its hydrating properties and electrolyte content.

Coconut Husk: This is the fibrous outer layer of the coconut that is removed to access the meat inside. It can be used to make a variety of products, including rope, mats, and fuel.

Coconut Shell: This is the hard, outer layer of the coconut that is removed to access the meat inside. It can be used to make charcoal, activated carbon, and other industrial products.

Glycerin: This is a byproduct of the chemical process used to refine coconut oil. It is commonly used in cosmetics, personal care products, and pharmaceuticals.

Overall, coconut oil production generates a range of co-products, which can have various applications across different industries. These co-products can add value to the coconut processing industry and contribute to more sustainable and efficient use of coconut resources.

Soy oil

As mentioned above, in the data that we currently have, we only have soy oil listed as a coproduct, since Rich doesn't include it as an ingredient. Since I'm including it as an ingredient, I will need to find the yield factors for the coproducts associated with soy oil.

In summary, basically we need to add rows for soy oil, coconut oil, and pea flour, and figure out their co products and associated yields. Do we want to use just global average yields? Or should I do what we've done in the past and figure out the top ten countries that use these ingredients and their associated yields for each ingredient? (I think that is what @cottrellr did?)

[gclawson1]

Coconut oil deep dive

https://ift.onlinelibrary.wiley.com/doi/10.1111/1750-3841.14223

Destabilization of emulsion in coconut milk brings about the collapse of the emulsion, from which virgin coconut oil (VCO) can be obtained. Yield, characteristics, and properties of VCO are governed by the processes used for destabilizing coconut milk. VCO is considered to be a functional oil and is rich in medium chain fatty acids with health advantages.

Well-known products of coconut palm include coconut oil, coconut milk, coconut water and coconut meat. Coconut milk is generally extracted from grated coconut meat after pressing or squeezing with or without the addition of water. Besides serving as a food ingredient, coconut milk is used for the production of virgin coconut oil (VCO), for which collapse of coconut milk emulsion is required. VCO is commonly manufactured from coconut meat (wet kernel) by natural or mechanical means without or with the application of heat.

Based on these passages, it seems like a safe bet for coconut oil co-products are coconut milk, coconut meat, and coconut water. Maybe coconut husk as well? (Are the co-products we are accounting for things only used for food?)

Asked chatgpt what the yield rates for coconut oil and its coproducts are:

Coconut Oil: The yield rate for coconut oil production can range from 50% to 70% of the total weight of the coconut meat used. This means that for every 100 kilograms of coconut meat, 50 to 70 kilograms of coconut oil can be produced.

Coconut Meal: The yield rate for coconut meal can range from 20% to 30% of the total weight of the coconut meat used. This means that for every 100 kilograms of coconut meat, 20 to 30 kilograms of coconut meal can be produced.

Coconut Milk: The yield rate for coconut milk can range from 10% to 15% of the total weight of the coconut meat used. This means that for every 100 kilograms of coconut meat, 10 to 15 liters of coconut milk can be produced.

Coconut Water: The yield rate for coconut water can range from 20% to 25% of the total weight of the coconut used. This means that for every 100 coconuts, 20 to 25 liters of coconut water can be produced.

Coconut Husk and Shell: The yield rate for coconut husk and shell can vary depending on the intended use. For example, the yield rate for coconut husk fiber can range from 30% to 40% of the total weight of the coconut husk, while the yield rate for coconut shell charcoal can range from 20% to 30% of the total weight of the coconut shell.

Check against the Technical Conversion Factors for Agricultural Commodities (not comprehensive):

Country	Food type	Extraction rate (%)
Ghana	Oil of coconuts	60
Kenya	Oil of coconuts	62
Mozambique	Oil of coconuts	64
Nigeria	Oil of coconuts	64
South Africa	Oil of coconuts	64
Tanzania	Oil of coconuts	60
Brazil	Oil of coconuts	56
India	Oil of coconuts	62
Indonesia	Oil of coconuts	54
Japan	Oil of coconuts	61
Phillippines	Oil of coconuts	64
Thailand	Oil of coconuts	64
Norway	Oil of coconuts	64
UK	Oil of coconuts	63
Australia	Oil of coconuts	59

Most countries report ~60% extraction rate for coconut oil. Assuming that "cake of coconuts" is equivalent to coconut meal, then nearly all are 30-35% extraction rates. Coconut milk and water are not reported in the TCFs.

On the extraction of coconut milk: https://www.sciencedirect.com/science/article/pii/S0308814606001634

Another paper on extraction of coconut milk: https://article.sciencepublishinggroup.com/html/10.11648.j.ijnfs.20160506.13.html

The result in Table 2 showed that the predicted and observed coconut milk yield varied from 53.49 to 61.84% and 53.33±0.33 to 61.67±0.30% respectively.

For coconut milk, getting some mixed results. One paper suggests ~30% yield extraction rates, the other ~55%.

Having a lot of trouble finding yield rates for coconut water... but managed to find this: https://www.researchgate.net/publication/346773633_Coconut_Water_-_A_Nature's_miracle_health_drink_Chemistry_Health_Benefits_Packaging_Storage_and_Technologies_A_Review

This suggests coconut water is on average 18% of the coconut.

[cottrellr]

Awesome. Yes great to look into the specifics of different ingredients. And we can slowly build a portfolio of different ingredients as this will be useful from a final app for Biomar perspective.

As we chatted about, I don't think we need to keep it to country specific yields for each ingredient as it gets unwieldy and is inconsistent for ingredients outside of the FAO technical conversion factors document. And from what I've seen in my analysis, it makes very little meaningful difference as most yields are within 2-5% of each other, and when combined with variation in coproduct yield, it often creates a similar raw material conversion factor 1.19 vs 1.21 for instance.

One thing that is important to note is I have seen through looking through the manufacturing processes for different ingredients that while literature may cite some products as coproducts of coconut processing - not all products listed as produced from the processing are coproducts of each other. I made this mistake with SPC and SPI and SBM leading to conversion factors that were too small. SPC is produced from the dehulled soybeans OR SPI is or SBM. So they are alternative products you can create - just flagging that understanding the manufacturing process helps with making sure this is right.

---

Richard S. Cottrell Research Fellow in Aquaculture Sustainability Institute for Marine and Antarctic Studies College of Sciences and Engineering University of Tasmania

Theme Co-Lead, Sustainable Futures and Planetary Health Centre for Marine Socioecology University of Tasmania

Size Ecology Labhttps://www.sizeecology.org/ | Centre for Marine Socioecologyhttps://marinesocioecology.org/themes/sustainable-futures-and-planetary-health/ Google Scholarhttps://scholar.google.com/citations?user=1pLCMKIAAAAJ&hl=en | ORCIDhttps://orcid.org/my-orcid?orcid=0000-0002-6499-7503 | @RichCottrell22https://twitter.com/RichCottrell22

On 10 May 2023, at 8:18 am, Gage Clawson @.**@.>> wrote:

Finally getting back into data prep!

We still have a couple of unresolved issues regarding yield conversion factors for allocation that I need. I'll be working this week to fix these.

1. Difference between pea starch and pea flour. The Aas et al. 2022 paper reports "pea flour", so that is the ingredient that I am using in my analysis. However, we currently only have yield data for pea starch.
2. Need to find yield info and coproduct yield info for coconut oil
3. Soy oil is only listed as a coproduct in the data that Rich uses. I'm using it as an ingredient in my analysis, so I need it to be an ingredient, and have co product yield factors (explained below).

Pea starch vs pea flour:

(I asked chatgpt for this description)

Pea starch is a fine, white powder that is extracted from peas through a process that involves washing, grinding, and separating the starch from the other components of the pea. Pea starch is a versatile ingredient that can be used in a wide range of food applications, including as a thickener, binder, and stabilizer. It is often used as a substitute for other starches, such as cornstarch or potato starch, in gluten-free and vegan recipes. Pea starch is also used in the production of processed foods, such as baked goods, snacks, and pasta.

Pea flour, on the other hand, is made by grinding whole peas into a fine powder. Unlike pea starch, pea flour contains not only the starch but also the protein, fiber, and other nutrients found in the pea. Pea flour is often used as a source of plant-based protein in a variety of food products, including bread, pasta, and snacks. It can also be used as a gluten-free flour substitute in baking, although its texture and flavor are different from wheat flour.

So pea starch is a material that is derived (or could be derived) from pea flour.

Coconut oil

We just don't have any information on this yet, so need to track it down!

Soy oil

As mentioned above, in the data that we currently have, we only have soy oil listed as a coproduct, since Rich doesn't include it as an ingredient. Since I'm including it as an ingredient, I will need to find the yield factors for the coproducts associated with soy oil.

In summary, basically we need to add rows for soy oil, coconut oil, and pea flour, and figure out their co products and associated yields. Do we want to use just global average yields? Or should I do what we've done in the past and figure out the top ten countries that use these ingredients and their associated yields for each ingredient? (I think that is what @cottrellrhttps://github.com/cottrellr did?)

image.png (view on web)https://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/assets/33332753/4ea0aead-8099-4b9b-b4ab-f4b3b7751c2f

— Reply to this email directly, view it on GitHubhttps://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/issues/4#issuecomment-1540960404, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AJK3YJA345FVRLZJP3LYFQLXFK7DFANCNFSM6AAAAAAWLHJU4U. You are receiving this because you were mentioned.Message ID: @.***>

This email is confidential, and is for the intended recipient only. Access, disclosure, copying, distribution, or reliance on any of it by anyone outside the intended recipient organisation is prohibited and may be a criminal offence. Please delete if obtained in error and email confirmation to the sender. The views expressed in this email are not necessarily the views of the University of Tasmania, unless clearly intended otherwise.

[cottrellr]

This is tricky with coconut oil.

So if you have X amount of coconut oil demand from Y amount of salmon production, I think we should to try to fuel that demand with the minimum amount of biomass necessary. So to me - this means that you use the meat to create milk and that milk to create the oil. So let’s assume that you use all the coconut meat for milk production (so meat is no longer a food option), and then all milk to create oil (so it’s no longer usable as milk for food either). So would this actually mean that the coproducts are water and husk and the meat waste from the pressing for milk and the waste from milk to oil process? This is where the manufacturing process should be clear to us.

---

Richard S. Cottrell Research Fellow in Aquaculture Sustainability Institute for Marine and Antarctic Studies College of Sciences and Engineering University of Tasmania

Theme Co-Lead, Sustainable Futures and Planetary Health Centre for Marine Socioecology University of Tasmania

Size Ecology Labhttps://www.sizeecology.org/ | Centre for Marine Socioecologyhttps://marinesocioecology.org/themes/sustainable-futures-and-planetary-health/ Google Scholarhttps://scholar.google.com/citations?user=1pLCMKIAAAAJ&hl=en | ORCIDhttps://orcid.org/my-orcid?orcid=0000-0002-6499-7503 | @RichCottrell22https://twitter.com/RichCottrell22

On 10 May 2023, at 10:26 am, Gage Clawson @.**@.>> wrote:

Coconut oil deep dive

https://ift.onlinelibrary.wiley.com/doi/10.1111/1750-3841.14223https://ift.onlinelibrary.wiley.com/doi/10.1111/1750-3841.14223

Destabilization of emulsion in coconut milk brings about the collapse of the emulsion, from which virgin coconut oil (VCO) can be obtained. Yield, characteristics, and properties of VCO are governed by the processes used for destabilizing coconut milk. VCO is considered to be a functional oil and is rich in medium chain fatty acids with health advantages.

Well-known products of coconut palm include coconut oil, coconut milk, coconut water and coconut meat. Coconut milk is generally extracted from grated coconut meat after pressing or squeezing with or without the addition of water. Besides serving as a food ingredient, coconut milk is used for the production of virgin coconut oil (VCO), for which collapse of coconut milk emulsion is required. VCO is commonly manufactured from coconut meat (wet kernel) by natural or mechanical means without or with the application of heat.

Based on these passages, it seems like a safe bet for coconut oil co-products are coconut milk, coconut meat, and coconut water. Maybe coconut husk as well? (Are the co-products we are accounting for things only used for food?)

Asked chatgpt what the yield rates for coconut oil and its coproducts are:

Coconut Oil: The yield rate for coconut oil production can range from 50% to 70% of the total weight of the coconut meat used. This means that for every 100 kilograms of coconut meat, 50 to 70 kilograms of coconut oil can be produced.

Coconut Meal: The yield rate for coconut meal can range from 20% to 30% of the total weight of the coconut meat used. This means that for every 100 kilograms of coconut meat, 20 to 30 kilograms of coconut meal can be produced.

Coconut Milk: The yield rate for coconut milk can range from 10% to 15% of the total weight of the coconut meat used. This means that for every 100 kilograms of coconut meat, 10 to 15 liters of coconut milk can be produced.

Coconut Water: The yield rate for coconut water can range from 20% to 25% of the total weight of the coconut used. This means that for every 100 coconuts, 20 to 25 liters of coconut water can be produced.

Coconut Husk and Shell: The yield rate for coconut husk and shell can vary depending on the intended use. For example, the yield rate for coconut husk fiber can range from 30% to 40% of the total weight of the coconut husk, while the yield rate for coconut shell charcoal can range from 20% to 30% of the total weight of the coconut shell.

Check against the Technical Conversion Factors for Agricultural Commodities (not comprehensive):

Country Food type Extraction rate (%) Ghana Oil of coconuts 60 Kenya Oil of coconuts 62 Mozambique Oil of coconuts 64 Nigeria Oil of coconuts 64 South Africa Oil of coconuts 64 Tanzania Oil of coconuts 60 Brazil Oil of coconuts 56 India Oil of coconuts 62 Indonesia Oil of coconuts 54 Japan Oil of coconuts 61 Phillippines Oil of coconuts 64 Thailand Oil of coconuts 64 Norway Oil of coconuts 64 UK Oil of coconuts 63 Australia Oil of coconuts 59

Most countries report ~60% extraction rate for coconut oil. Assuming that "cake of coconuts" is equivalent to coconut meal, then nearly all are 30-35% extraction rates. Coconut milk and water are not reported in the TCFs.

On the extraction of coconut milk: https://www.sciencedirect.com/science/article/pii/S0308814606001634https://www.sciencedirect.com/science/article/pii/S0308814606001634

image.png (view on web)https://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/assets/33332753/993455b0-baaa-45e3-979b-e3f269ff6143

image.png (view on web)https://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/assets/33332753/577d7953-b1dd-44ab-9908-97629c6517b8

Another paper on extraction of coconut milk: https://article.sciencepublishinggroup.com/html/10.11648.j.ijnfs.20160506.13.htmlhttps://article.sciencepublishinggroup.com/html/10.11648.j.ijnfs.20160506.13.html

The result in Table 2 showed that the predicted and observed coconut milk yield varied from 53.49 to 61.84% and 53.33±0.33 to 61.67±0.30% respectively.

For coconut milk, getting some mixed results. One paper suggests ~30% yield extraction rates, the other ~55%.

Having a lot of trouble finding yield rates for coconut water... but managed to find this: https://www.researchgate.net/publication/346773633_Coconut_Water_-_A_Nature's_miracle_health_drink_Chemistry_Health_Benefits_Packaging_Storage_and_Technologies_A_Reviewhttps://www.researchgate.net/publication/346773633_Coconut_Water_-_A_Nature's_miracle_health_drink_Chemistry_Health_Benefits_Packaging_Storage_and_Technologies_A_Review

image.png (view on web)https://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/assets/33332753/fe61c0d6-134e-417c-b181-77dba849a1ab

This suggests coconut water is on average 18% of the coconut.

— Reply to this email directly, view it on GitHubhttps://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/issues/4#issuecomment-1541055248, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AJK3YJCPKIMWWKJFKG237MDXFLODRANCNFSM6AAAAAAWLHJU4U. You are receiving this because you were mentioned.Message ID: @.***>

This email is confidential, and is for the intended recipient only. Access, disclosure, copying, distribution, or reliance on any of it by anyone outside the intended recipient organisation is prohibited and may be a criminal offence. Please delete if obtained in error and email confirmation to the sender. The views expressed in this email are not necessarily the views of the University of Tasmania, unless clearly intended otherwise.

[cottrellr]

I think the oil extraction rates in the technical conversion factor seem super high - around %60 seems implausible if water is 18% (seems reasonable as someone who has drunk his body weight in water from directly from coconuts) and husk is 30% (which again is reasonable on average - coconuts are almost all husk) - neither of which you get oil from. the 60% for coconut oil I assume would be for a different primary material (I'm thinking dehusked coconuts perhaps - the ones you knock off the top of posts at the fair lol)

[gclawson1]

You're right.. the oil rate does seem high. I'm guessing that the extraction rate is referring to the yield of oil from the coconut meat, rather than the whole coconut. So if the coconut meat yield (lets assume that is the kernel rate) is ~30%, then we want 60% of 30%? Which would be 18%?

[gclawson1]

Found this report on the processing of coconut oil: https://www.ofimagazine.com/content-images/news/Coconut_oil_processing.pdf

"Coconut oil is derived from copra, which the dried kernel or 'meat' of coconut".

Processing flow of coconut oil:

Steps for regular coconut oil:

1. De-hulling: "Involves cracking the shell to take out the meat or kernel.
2. Drying the kernel: "The kernel contains about 50% moisture, and it is dried to a moisture content of 6-8% before oil extraction.
3. Oil extraction: "The dried kernel has an oil content of about 64%. Usually the oil is extracted from the copra by pressing in screw presses (expellers), followed by solvent extraction to recover the residual oil from the cake."
4. Refining to make safe for human consumption and better shelf life.

Steps for virgin coconut oil: VCO is extracted from coconut milk obtained from the mature kernel of a coconut.

1. De-husk
2. De-shell
3. Blanching
4. Draining
5. Grinding the meat
6. Press out the milk from the meat
7. Separate the oil from the milk
8. Filter

Some notes on possible by-products:

Coconut flour

"Coconut milk is a natural oil-in-water emulsion. The oil separated by centrifugation is filtered to remove any solids present. The residue, flake/ defatted desiccated coconut is dried and is often used as flour."

[Residual cake}(http://www.agritech.tnau.ac.in/expert_system/coconut/coconut/coconut_processing.html#:~:text=The%20residual%20cake%20after%20the,fats%20for%20baking%20and%20confectionery.) or copra meal:

"The residual cake after the oil has been extracted is used as cattle food." "The main coconut by-product is the copra meal. Depending on the oil extraction method, the oil residue in the marketed product ranges from 1% to 22% (Göhl, 1982). The terms “copra cake” and “copra meal” sometimes refer respectively to the mechanically extracted and the solvent extracted product (FAO/IAEA, 2001). However, the names are often interchangeable in practice and in this datasheet copra meal will be used as a generic term to designate the oil by-product."

Coconut water: Based on feedipedia, we don't want to use coconut water as co-product: "coconut water is usually wasted when the nuts are split open. The dry matter content of coconut water declines as the nut matures and is a meagre source of nutrients when the nuts are harvested for copra (Göhl, 1982)."

Summary:

Based on the above, I think the TCF of 60% refers to 60% of the kernel. If the meat (kernel) has a yield of ~30%, then the oil would be 60% of 30% = 18%?.

I think it is safe to assume possible co-products could be coconut flour (if we consider human consumption) and residual cake (copra meal) for cattle feed. Feedipia indicates that coconut meal is a by-product of oil extraction from dried coconut kernels.

For now, I'll just assume that the only by-product of coconut oil is the coconut meal for cattle feed. For now I'll assume that the extraction rate for coconut meal is the TCF "cake of coconuts" yield, ~30%. I'll also assume that the coconut oil rate is 18%.

[gclawson1]

Pea flour deep dive

From wikipedia:

The cleaning process: Cleaning uses equipment such as indent cleaners, which are used to remove impurities. This process ensures that any allergens, such as wheat, barley and other seeds are removed, as these products contain gluten.[9] If not removed, it would affect its classification as a gluten-free product.[9]

The splitting process: Following cleaning, the pea is split and "dehulled" using a dehuller.[9] The dehullers are a device that splits the pea seed and extracts the part of the whole seed which is the hull.[9]

Further processing: Following the splitting process, the split peas are further processed either into starch, protein, and flour fractions.[9]

This makes it sound like pea flour, starch, and protein can be co-products from the same manufacturing process.

We already have yield rates for pea protein concentrate (0.23) and pea starch (0.5). I'm guessing that pea flour would be higher than that as pea protein concentrate and pea starch are both derived from pea flour.

0.23 and 0.5 are backed up from this article: https://www.sciencedirect.com/science/article/pii/S0308814618317680?via%3Dihub

Field pea (Pisum sativum L.) is one of the major legumes which have been subjected to dry fractionation because of its nutritional and functional properties (Pelgrom, Boom, & Schutyser, 2015a). It typically contains 23% protein (dry base), 48% starch (dry base), 4% total lipids, 3% ash, 8% sugar, etc. (Wu & Nichols, 2005)

This paper shows milling yields for different types of pulses. Indicates ~85% yield for yellow peas:

[gclawson1]

You're right.. the oil rate does seem high. I'm guessing that the extraction rate is referring to the yield of oil from the coconut meat, rather than the whole coconut. So if the coconut meat yield (lets assume that is the kernel rate) is ~30%, then we want 60% of 30%? Which would be 18%?

Actually, the TCF lists "Copra" which is just coconut meal. So I will use the copra value instead of the "Cake of coconuts" value. So the oil yield rate will actually be (for Indonesia as example), 54% (the oil yield listed) of 20% (the copra yield listed) = 10.8%.

[cottrellr]

Awesome, this makes much more sense. 18% looks much more reasonable for oil. In the TCF the origin raw material from which the extraction happened isn’t always clear - this is something I will have to check with my conversion factors.

---

Richard S. Cottrell Research Fellow in Aquaculture Sustainability Institute for Marine and Antarctic Studies College of Sciences and Engineering University of Tasmania

Theme Co-Lead, Sustainable Futures and Planetary Health Centre for Marine Socioecology University of Tasmania

Size Ecology Labhttps://www.sizeecology.org/ | Centre for Marine Socioecologyhttps://marinesocioecology.org/themes/sustainable-futures-and-planetary-health/ Google Scholarhttps://scholar.google.com/citations?user=1pLCMKIAAAAJ&hl=en | ORCIDhttps://orcid.org/my-orcid?orcid=0000-0002-6499-7503 | @RichCottrell22https://twitter.com/RichCottrell22

On 11 May 2023, at 4:40 am, Gage Clawson @.**@.>> wrote:

Found this report on the processing of coconut oil: https://www.ofimagazine.com/content-images/news/Coconut_oil_processing.pdfhttps://www.ofimagazine.com/content-images/news/Coconut_oil_processing.pdf

"Coconut oil is derived from copra, which the dried kernel or 'meat' of coconut".

Processing flow of coconut oil:

Steps for regular coconut oil:

1. De-hulling: "Involves cracking the shell to take out the meat or kernel.
2. Drying the kernel: "The kernel contains about 50% moisture, and it is dried to a moisture content of 6-8% before oil extraction.
3. Oil extraction: "The dried kernel has an oil content of about 64%. Usually the oil is extracted from the copra by pressing in screw presses (expellers), followed by solvent extraction to recover the residual oil from the cake."
4. Refining to make safe for human consumption and better shelf life.

Steps for virgin coconut oil: VCO is extracted from coconut milk obtained from the mature kernel of a coconut.

1. De-husk
2. De-shell
3. Blanching
4. Draining
5. Grinding the meat
6. Press out the milk from the meat
7. Separate the oil from the milk
8. Filter

Some notes on possible by-products:

Coconut flour

"Coconut milk is a natural oil-in-water emulsion. The oil separated by centrifugation is filtered to remove any solids present. The residue, flake/ defatted desiccated coconut is dried and is often used as flour."

[Residual cake}(http://www.agritech.tnau.ac.in/expert_system/coconut/coconut/coconut_processing.html#:~:text=The%20residual%20cake%20after%20thehttp://www.agritech.tnau.ac.in/expert_system/coconut/coconut/coconut_processing.html#:~:text=The%20residual%20cake%20after%20the,fats%20for%20baking%20and%20confectionery.) or copra mealhttps://www.feedipedia.org/node/46:

"The residual cake after the oil has been extracted is used as cattle food." "The main coconut by-product is the copra meal. Depending on the oil extraction method, the oil residue in the marketed product ranges from 1% to 22% (Göhl, 1982https://www.feedipedia.org/node/1661). The terms “copra cake” and “copra meal” sometimes refer respectively to the mechanically extracted and the solvent extracted product (FAO/IAEA, 2001https://www.feedipedia.org/node/5821). However, the names are often interchangeable in practice and in this datasheet copra meal will be used as a generic term to designate the oil by-product."

Coconut water: Based on feedipediahttps://www.feedipedia.org/node/46, we don't want to use coconut water as co-product: "coconut water is usually wasted when the nuts are split open. The dry matter content of coconut water declines as the nut matures and is a meagre source of nutrients when the nuts are harvested for copra (Göhl, 1982https://www.feedipedia.org/node/1661)."

Summary:

Based on the above, I think the TCF of 60% refers to 60% of the kernel. If the meat has a yield of 30%, then the oil would be 60% of 30% = 18%?.

I think it is safe to assume possible co-products could be coconut flour (if we consider human consumption) and residual cake (copra meal) for cattle feed. Feedipiahttps://www.feedipedia.org/node/46 indicates that coconut meal is a by-product of oil extraction from dried coconut kernels.

For now, I'll just assume that the only by-product of coconut oil is the coconut meal for cattle feed. For now I'll assume that the extraction rate for coconut meal is the TCF "cake of coconuts" yield, ~30%.

— Reply to this email directly, view it on GitHubhttps://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/issues/4#issuecomment-1542643580, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AJK3YJE666STN4GP2FMEL5LXFPOK7ANCNFSM6AAAAAAWLHJU4U. You are receiving this because you were mentioned.Message ID: @.***>

This email is confidential, and is for the intended recipient only. Access, disclosure, copying, distribution, or reliance on any of it by anyone outside the intended recipient organisation is prohibited and may be a criminal offence. Please delete if obtained in error and email confirmation to the sender. The views expressed in this email are not necessarily the views of the University of Tasmania, unless clearly intended otherwise.

[cottrellr]

To answer your question re soy oil - I think at a bear minimum the coproducts of soy oil is lecithin and defatted soybean flakes - this is an figure from van den Berg et al 2022

And this report mirrors that - note SBM, SPC, SPI all have soy oil and lecithin as coproducts in this but soy oil and lecithin don't necessarily have SPC, SBM, or SPI as coproducts because you don't necessarily produce these from defatted flakes (you might just use the defatted flakes).

[gclawson1]

From what I've gathered, defatted flakes are always used as soybean meal, functionally being the same thing. They can be produced into SPC as well. So I think that the soy oil co-product should be soybean meal, rather than "defatted flakes". This diagram makes it seem like there doesn't have to be any real processing to go from defatted flakes to SBM, which would justify this assumption:

[gclawson1]

the commodity tree from the TCF makes me think the same thing:

[gclawson1]

Finished putting together the top 10 countries and their yield extraction rates. I checked over the numbers already in there, and everything checked out. I started a new folder in the onedrive called "allocation_new". https://universitytasmania.sharepoint.com/:f:/r/sites/IMASSustainableAquafeedsProject-Feedmapping/Shared%20Documents/Feed%20mapping/data/[allocation_new](https://universitytasmania.sharepoint.com/:f:/r/sites/IMASSustainableAquafeedsProject-Feedmapping/Shared%20Documents/Feed%20mapping/data/allocation_new?csf=1&web=1&e=72Z50m)?csf=1&web=1&e=72Z50m

I'm gonna take an average of these yield rates, and will produce a data sheet with one yield factor for each product and co-product. I'll also link it to the existing gross energy values for energetic allocation.

[cottrellr]

Superb work, Gage. Happy with how you present the SBM/defatted flakes distinction too. We can include this information in the manufacturing process doc.

Sent from my iPhone

On 12 May 2023, at 5:21 am, Gage Clawson @.***> wrote:



Finished putting together the top 10 countries and their yield extraction rates. I checked over the numbers already in there, and everything checked out. I started a new folder in the onedrive called "allocation_new". https://universitytasmania.sharepoint.com/:f:/r/sites/IMASSustainableAquafeedsProject-Feedmapping/Shared%20Documents/Feed%20mapping/data/[allocation_new](https://universitytasmania.sharepoint.com/:f:/r/sites/IMASSustainableAquafeedsProject-Feedmapping/Shared%20Documents/Feed%20mapping/data/allocation_new?csf=1&web=1&e=72Z50m)?csf=1&web=1&e=72Z50mhttps://universitytasmania.sharepoint.com/:f:/r/sites/IMASSustainableAquafeedsProject-Feedmapping/Shared%20Documents/Feed%20mapping/data/%5Ballocation_new%5D(https://universitytasmania.sharepoint.com/:f:/r/sites/IMASSustainableAquafeedsProject-Feedmapping/Shared%20Documents/Feed%20mapping/data/allocation_new?csf=1&web=1&e=72Z50m)?csf=1&web=1&e=72Z50m

I'm gonna take an average of these yield rates, and will produce a data sheet with one yield factor for each product and co-product. I'll also link it to the existing gross energy values for energetic allocation.

— Reply to this email directly, view it on GitHubhttps://github.com/Sustainable-Aquafeeds-Project/feed_biodiv_impact_mapping/issues/4#issuecomment-1544552688, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AJK3YJBA3TAXCRRELACQT2DXFU323ANCNFSM6AAAAAAWLHJU4U. You are receiving this because you were mentioned.Message ID: @.***>

This email is confidential, and is for the intended recipient only. Access, disclosure, copying, distribution, or reliance on any of it by anyone outside the intended recipient organisation is prohibited and may be a criminal offence. Please delete if obtained in error and email confirmation to the sender. The views expressed in this email are not necessarily the views of the University of Tasmania, unless clearly intended otherwise.

[gclawson1]

Calculated average values for all yield data, and found gross energy values for coconut and copra meal. Joined them together in one table and put onto teams in allocation_raw. Spreadsheet is called "crop_ingredient_allocation_factors.csv":

[gclawson1]

Also have this in a script that should be compatible in the pressures project. The FAO item names might be different, since I'm using the newer version of the crop production, but the item codes are the same: feed_biodiv_impact_mapping/prep/00a_crop_raw_materials_conversions.Rmd

[gclawson1]

A few weeks ago we discussed including other aquafeed ingredients than just salmon feed ingredients for our processing flow and conversion factor data paper.

Rich pointed me to a paper that had feed compositions of freshwater fish, diadromous fish, marine fish, and crustaceans: https://www-sciencedirect-com.proxy.library.ucsb.edu/science/article/pii/S0048969715304666#s0065

Their supplementary table 3 has all of the ingredients they considered listed. I've put together a list of those ingredients, compared it to the ingredients we already have:

Pahlow Table S3	Currently included in our data	Notes
Rendered animal products (carcass meal and milk powder)
Barley
Broad bean meal	x	Assuming same as faba beans
Canola meal	x
Canola oilcake	x
Cassava meal	x
Coconut meal (copra)	x	We named it "cake of coconut"
Corn	x
Corn meal
Corn gluten meal	x
Cottonseed meal
Cottonseed oilcake
Faba bean meal	x
Field pea meal	x	Assuming same as pea flour (e.g., "peasemeal")
Groundnut meal
Groundnut cake
Lupin kernel meal
Mustard meal
Mustard oil cake
Rapeseed cake	x	Assuming same as meal
Rapeseed meal	x
Rice
Rice bran
Rice, broken
Rice bran (de-oiled)
Rice grain ground
Rice groats and meal
Sesame seed cake
Sugar cane molassess
Soybean meal	x
Soybean cake	x	Assuming same as meal
Soy protein concentrate	x
Sorghum
Sunflower meal	x
Sunflower oil cake	x	Assuming same as meal
Wheat	x
Wheat bran
Wheat gluten meal	x
Wheat meal, wheat flour, ground wheat		Unsure on this one, would this just be considered wheat? Since it is just ground.
Corn oil
Cottonseed oil
Fish oil	x
Groundnut oil
Palm oil
Rapeseed oil	x
Soybean oil	x
Sunflower oil	x
Corn starch	x
Cassava starch
Potato starch
Wheat starch	x
Cellulose

There are 27 ingredients listed above that we do not account for, however, we probably don't need to include all of those, like some of the rice ones, cellulose, etc, that might be very uncommon.

Additionally, we have some ingredients that they do not include:

maize steep liquor corn gluten feed pea protein concentrate pea pulp pea solubles linseed meal linseed oil guar guar guar gum powder pea flour

I'll start researching some of the more obvious ones we want to include (like cassava meal/starch, cottonseed, corn meal/oil) and add them to the paper and extraction rates.

Also, I think it might be a good idea to just create a repo specifically for the processing flows and conversion factors. We could put all of the data and scripts for the conversion factor calculations and any figures in it.