My notes and musings …

Simple quality tests for small millet rice

with one comment

A couple of weeks ago, I started sharing some of my understandings of millets, their nutrition, economics, ecology and technology in 2 to 3 hr sessions. The first of these was on Simple quality tests for staple forms of millets. For convenience and brevity, I have split it into three different sets – one on millet rice, another on millets grits and rava and a third one on millet flours and mixes. Here’s the first one on millet rices …


Written by Dwiji

Saturday, October 22nd, 2016 at 22:37

Posted in Food, Millets, Uncategorized

Tagged with , ,

Costing whole grain millet rice based diet …

with one comment

When I talk about millets with those exploring better food options, a common concern expressed is the cost factor. “I am totally convinced from nutritional, environmental and social justice perspectives, but it is too costly. Not affordable.” It is a valid concern. Lets crunch the numbers on this …

Millet rices have a high satiety index, especially whole grain (unpolished) millet rices (wgmr). One feels more full on eating a smaller quantity of cooked wgmr than when eating the same quantity of polished paddy rice (ppr), or even whole grain (unpolished) paddy rice for that matter. But for now, lets look at how much it costs to serve a wgmr based meal as compared to a ppr based one.

Kodo millet rice, cooked using a pressure cooker.

Kodo millet rice, cooked using a pressure cooker.

In a meal, I eat about 300 to 400 gms of cooked wgmr. When I have a ppr based meal, I eat 400 to 600 gms before I feel full. Millets are thirsty grains and cooking wgmr takes more water than cooking ppr – 1:3 for wgmr as compared to 1:2 for ppr. So for a meal, I need at the most 400gms/4=100 gms of wgmr (dry, uncooked) and 600gms/3=200 gms of ppr  (dry, uncooked). In BLR good quality wgmr will cost me about Rs. 100/kg and decent ppr would be about Rs.40/kg. So my cost per meal for wgmr is therefore Rs. 100*(100/1000)=Rs. 10 as compared to Rs. 40*(200/1000)=Rs. 8 for ppr.

This is just from sheer quantity and cost perspective. Factoring in the difference in nutrition that the two meals would provide, the cost of wgmr is already justified. But wait, there is yet another factor that needs to be taken into account.

Again, purely from the quantity perspective, almost everyone who has eaten wgmr based meals share this experience that even after 4 hours of the meal, they typically do not feel hungry for a snack. The same individuals after eating a ppr based meal would be seeking out something to snack on as they touch the 3 hours mark since their meal. This is primarily because of the high fibre (and small carbohydrate to fibre ratio) of wgmr. But from a purely financial perspective, just the ingredients cost of a snack would be anywhere from Rs. 3 to 5. Adding this to the above, when we have a wgmr based meals, I would incur an expense of Rs.10 before I get to the next meal, while on days I eat ppr based meals, I would incur an expense of Rs. 11 before I get to the next meal.


Please see some tips on cooking wgmr here to improve the experience of cooking and eating wgmr based dishes. And just search for ‘millets buy the-place-where-you-live’ online and you will find a whole list of online and real world stores where you can buy millets. Do call and check with them if the millet rices they sell are the whole grain (unpolished) kind before you decide to make the purchase.

Whole grain (unpolished) millet rices are not just good for our health, they are good for our pocket books too !!

Written by Dwiji

Saturday, October 22nd, 2016 at 06:17

Posted in Food, Millets

Parameters to compare different millet grain hulling machines

with 2 comments

This is a Q that keeps coming up again and again. And given how much impact this has on realizing more millet processing units, I am kicking myself for not having posted this earlier.
There are many individuals and organizations thinking of starting millet processing. And everyone arrives at a realization that the biggest challenge is in getting a good dehusking machine. This is only part of the challenge, but thats for another day. So how does one evaluate the performance of a dehusking / hulling machine? Broadly, we can look at these parameters under two categories.
What does it take to get it to work?
These are reasonably self evident and do not need too much explaining.
  1. How much electric power will it require to run?
  2. Does it require 3 phase or single phase electricity?
  3. What is the minimum quantity needed to get decent output?
  4. How many people are needed to get the design output?
  5. How much floor space does it require?
  6. Does it require an elevator to load grains into the machine?
  7. How many times do the grains need to be processed before we get desired quality output?
When working, what does it deliver & how much of it?

There are four parameters to quantify this:

  1. hulling efficiency: how many unhulled grains come through in 100 grains of the output after one pass through the huller.
  2. rice recovery percentage: weight of the rice fraction in the output as a percentage of input material.
  3. grain shattering percentage: the weight of the shattered grains in the output as a percentage of the input material
  4. through put / capacity: What is the maximum quantity of input grains it can process in an hour?
  5. bran loss & damage to bran estimation: how white are the millet rice kernels in the output? And what is the estimate on the extent of damage to the bran layer?
Note that the unhulled grains and the broken/shattered rice kernels in the output should be separated before weighing for the rice fraction or the shattered grains fraction of the output.
Do ask the machine supplier these questions / for the values of these parameters. This information could be in the product brochure or as part of a third party evaluation report. If you are looking to process a particular millet, make sure the numbers they share are for that millet. Yes, millets are rain fed crops and there will be variations based on variety, season, etc. But having these numbers and estimates is essential to understand how well the machine performs and what has gone into designing (and refining) it.
If you are looking for a machine to process multiple millets, please ask them to share information for at least three of them – either foxtail or little millet, kodo millet, and either proso millet or barnyard millet.

Written by Dwiji

Wednesday, September 28th, 2016 at 00:02

Immature millet grains – why remove when processing?

with 6 comments

The primary motivation to eat cereal grains such as paddy rice, millet rices, or wheat is the nutrition we can derive from these. The largest component, in terms of weight composition of the grain are the carbohydrates in each of these grains. The other nutritional components we can derive from these grains are fibre, minerals and essential fatty acids – to varying degrees depending on the nutritional content of individual grains. In a recent post, I had mentioned about how we can use the carbohydrate to fibre ratio as a fairly good indicator to identify a grain that suits one’s dietary needs.

Note the immature grains in this foxtail millet sample. They are the thin ones distributed throughout, some with a greenish color.

Note the immature grains in this foxtail millet sample. They are the thin ones distributed throughout, some with a greenish color.

When a grain is very light, it is not filled with enough carbohydrates in its endosperm – the hard part of the grain. These grains typically do not get dehusked properly during the hulling process. And even when they do, the millet rice kernel tend to shatter resulting in an increase in the grits among the millet rice kernels. These immature grains would also not taste good when eaten primarily due to the ill-formed starch component in the endosperm or the heart of the grain. So the cooking quality deteriorates dramatically even if we are able to process them to rice or rawa form.

The maturity of the oils – the fatty acids in the bran layer in such grains is also very low. This means that the oils go rancid very quickly in such immature grains even if

one is able to get the husk off without damaging the millet rice kernel. And once the oil on a few grains go rancid, it gives the entire package a foul odour and the rancidity spreads to the other mature grains too.

To summarize, removing the immature millet grains from the better formed ones during processing for the millet rice, improves  (i) the taste (ii) the cooking quality (iii) shelf life and (iv) the cleanliness of the product. Once separated, the light grains can be used for cattle feed as it is rich in cellulosic material.

Written by Dwiji

Tuesday, September 27th, 2016 at 22:47

How does eating millets help those with diabetes?

with 4 comments

Wheat, paddy and most millets have comparable glycemic index. i.e. the total quantity of sugar released in ones blood on eating 100gms of the grain, a direct function of carbohydrate content. Looking at the nutrition chart we can see that the carbohydrate content in all these grains are not too different.

As I had written about  earlier in the year, one needs to eat a lesser quantity of millets to feel as full as one would after eating another cereal grain such as polished paddy rice. So the serving size of millet based dishes are smaller and hence the glycemic load is lesser compared to preparations of polished paddy rice or refined flour.

And then there is the whole slow release aspect I had written about and identified how the carbohydrate to fibre ratio is a better indicator of this feature.

To sum up, after eating a millet based meal, the total sugar released into ones blood stream is reduced and the rate of increase initially, and the decrease later, in the blood sugar levels happens at a much more gradual rate when compared what is experienced after eating a meal of polished paddy rice or refined wheat flour.
Both these aspects are very beneficial for those with type 2 diabetes. Please note that diversity is almost always a good thing. It is advisable to include the various millets available in the local markets, unpolished/semi polished paddy rice and whole wheat flour to one’s diet.

Written by Dwiji

Friday, September 23rd, 2016 at 06:02

Will there be enough millets to feed the world?

with 3 comments

In my post a few days ago, I had presented my arguments as to why we should start including millets in our diets. Almost everyone who works on agriculture and food systems recognizes and acknowledges  that given the changing climate and rainfall patterns, in the not too distant future, we will not have sufficient paddy rice and wheat to feed the world if we continue down the track we are on. There are quite a few solutions that people are working on to mitigate this impending disaster. And from many perspectives, bringing millets back into our diets as a staple grain is the best way forward. But there are repeated Qs on whether we will have enough grains to feed the world if the whole world started eating millets.

We can answer our primary question by considering the following two Qs:

  1. what quantity of millet rice needs to be consumed to provide the same nutrition and make a person feel full after eating millets instead of paddy rice?
  2. on how much land can millets be grown on as compared to paddy?

The first Q has been answered in another recent post I wrote a few months ago on satiety index of millets. So, millet rice(s) adding up to  about two thirds the quantity of paddy rice would be sufficient to feed as many people as are surviving on paddy rice.

According to the Ministry of Agriculture, dept. of economics and stats, the average yield of paddy is about 2.2 tons / hectare. We do not have similar statistics for millets. But from experience, we can say that at the bare minimum, the average yield for millets is about 800 to a 1000 kgs/hectare.  Lets take the more conservative estimate (0.8 tons/hec) and to keep the math simple lets round this ratio to a third.

So taking into account the answer to one and these yield numbers, we end up with needing millets to be cultivated over at least twice the area as paddy is currently being cultivated over.

Even after the many thousands of crores that India has invested in irrigation, the land area on which paddy can be cultivated is less than 20% of the total cultivable area. Given how little millets demand of the soil, finding twice the area under paddy cultivation to cultivate millets on, is very much achievable.

So even with existing technology (seeds, practices, etc.), cultivating millets we can meet the food and nutritional requirements that paddy currently provides. Please note that I am not making the case for replacing paddy completely with millets, either at a food systems or at an individual level. I have considered the extreme situation to state my case.

Once we recognize that bringing back millets on the farms and in our diets is inevitable, a lot of smart minds can and will be brought in to work on millet cultivation and processing. Hopefully the lessons  learned from the green revolution will not be forgotten and we shall be smart enough to not fall into the same trap of ‘increase yield at all costs’  approach. Time will tell.

We do have a personal role to play in how this pans out. We must insist on policies that encourage sustainable farming methods and disincentivize chemical and energy intensive unsustainable practices.

Written by Dwiji

Wednesday, September 21st, 2016 at 06:50

Glycemic Index, Glycemic Load and the changes in blood sugar levels

with 8 comments

Glycemic Index is a measure of how much the blood sugar (i.e. glucose) increases after one eats a certain food. Pure glucose has a GI of 100 and all other foods are ranked with reference to this. GI is a relative ranking of different foods and is useful in estimating how much sugar will the body have to deal with. There are a lot of junk articles online about GI, it is always a good idea to go to the research programs rather than use popular media sources for information. For example, the University of Sydney’s website on GI is a good source of information on GI.

Glycemic Load is another number that factors in the recommended serving size. It is calculated by multiplying the GI with the recommended serving size. So it gives a much more realistic representation of how much increase in blood sugar one can expect after eating a certain food item. A high GI food item eaten in a small quantity would not increase the blood sugar level too much, whereas, a medium GI food eaten in large quantities during a meal would dump a whole lot more sugar than one would like.

Now one thing that neither of these two numbers capture is how soon the carbohydrates from the food reach the blood and increase blood sugar levels.

If one plots the blood sugar levels measured in an individual after eating a meal, it will increase up to a certain level and then begin to drop. GI is the area under the curve. Anyone who has studied college math (and remembers it!) will immediately recognize that the area can be the same for two curves with very different rate of rise, or even different peak values. When we are trying to identify how our body is stressed during digestion and absorption of nutrients, the rate of change in blood sugar levels is a critical parameter.

And surprisingly, there is a simple number one can use to get a measure of this – the carbohydrates to fibre ratio. For convenience, lets call it the ‘C/F’ of a food. Digestive juices need to soften and move through the ingested fibre before they can get to the sugars. So a high fibre food item would have a slower release of sugars compared to the same item with low fibre content. Based on this idea, the Harvard School of Public Health has promoted the 10:1 thumb rule to assess if a food item is whole grain or not.

So lets see the comparative nutritional table for millets and do the math to calculate the C/F for each of the food items. And the resulting table is …

Carbohydrate to Fibre ratio of different millets compared to other cereal grains

Carbohydrate to Fibre ratio of different millets compared to other cereal grains

There are two main advantages of using C/F rather than GI when assessing how good a particular food item is.

  1. It can be identified with just a simple division operation that a consumer can do looking at the nutritional label and does not need lab testing and trials.
  2. It is a better reflection of how it affects a person concerned about diabetes and controlling blood sugar levels.

The primary nutrition we derive from our cereals is the energy from its carbohydrate content. But this needs to be moderated to decrease its rate of release into the blood stream. And this is achieved by having a good fibre content. The secondary nutritional components we derive are the minerals and essential fatty acids. Note that within a grain, carbohydrates are found in the endosperm, the inner (typically white) portion of the grain that forms a majority of its volume and mass. Fibre is mostly found in the bran, a thin protective layer around the grain also rich in minerals, essential fatty acids.

So looking at it beyond the labels and nutritional analysis, it is a fairly simple thing to remember when shopping for something to eat:

Has bran? Good

No bran? Naaa!

I do advocate for people to move to whole grains, not just in millets, but also with paddy rice and wheat. And at the same time I recognize that as with everything else in life, one has to compromise  when negotiating what food to eat. So if eating polished/semi polished millets will get your family to accept it. Do make the shift. Gradually shift towards products that still have the bran layer.

Written by Dwiji

Tuesday, September 13th, 2016 at 06:52

%d bloggers like this: