A First Look at Flour

I want to spend some time looking at flour - how to measure it reproducibly and how different flours affect the way cookies turn out. I have a plan for that, and to start I went to the nearest large supermarket and bought eight bags of flour, each a different brand. That supermarket always has a wide choice of everything, and the choice of flours is just as amazing as everything else. (Way to go, Woodman's!) I bought every sort of flour that seemed appropriate for cookies, but skipped over whole wheat, bread, and cake flours, hence the paltry choice of a mere eight. Basically, I bought a bag of each flour that was designated "All-Purpose."

I have a few cookie recipes that use whole wheat, bread, or cake flour, but they probably number only fifteen recipes out of about 1800 in my cookbooks. I think eight kinds of flour is probably enough to look at for now, and so I made my cut-off there.

I was surprised at the price range, which went from $0.32 up to $1.25 per pound. The flours I bought and their per-pound prices were:

Shurfine All-Purpose Flour
$0.32
Pillsbury Best Unbleached All Purpose Flour
$0.38
Pillsbury Best All Purpose Flour
$0.38
Gold Medal All-Purpose Flour
$0.40
Dakota Maid All-Purpose Flour (milled by North Dakota Mill)
$0.44
Robin Hood Premium All-Purpose Flour
$0.54
Hodgson Mill Premium Naturally White Flour
$0.80
King Arthur 100% Organic All-Purpose Flour
$1.25

These were all 5-pound bags except for the King Arthur brand, which was only available off the shelf in 2-pound bags. Smaller quantities are usually more expensive, but I'm guessing it's the 100% organic label and what it implies that pushes the price of King Arthur flour way up. I wouldn't want to pay $6.25 per 5-pound bag of flour on a regular basis, though, that's for sure! But I'll do testing with it.

The first thing I did, just for fun, was to weigh each bag. Gave me a chance to use my new kitchen scale! :-) They're all 5-pound bags, supposedly (except for the King Arthur). But how much do they actually weigh? Five pounds is 2.268 kg, and all the bags were within 1% of the advertised weight except for the Robin Hood flour, which was 1.5% light. I wouldn't get upset about 1.5% light. I'm certain the weight of flour varies according to the moisture it contains at any given moment in time. I imagine it was almost exactly 5-pounds at the time it was packaged and some of the moisture dissipated since then.

There are certain obvious properties of flour which vary depending on the brand, and which are specified on the bag. The flours I bought displayed appellations of bleached, unbleached, enriched, and unbromated. Regarding the last property, the opposite is bromated flour. Some states require the manufacturer to state if the flour is bromated, and some don't. Thus you can't always tell from the bag if the flour is bromated or not. If it doesn't say, you have to ask the manufacturer. I'll be a little surprised if any of these affect the quality of cookies to a degree I can detect, but I suppose they could.

All additives are potential health issues unless they have been proven safe. The FDA approves the levels of additives in the flours manufactured today, but some people worry about it anyway. Unbleached, unbromated flour eliminates the potential health issue people get concerned about, chemical additives. On the other hand, bleached flour is whiter and affects the color of the product. I'm sure we've all been eating it since we were little and I haven't heard anything about evidence it is harmful. I'm not concerned about it myself, but everybody is free to choose whatever flour they want, and if someone is concerned about it, it might not bea bad idea to use unbleached unbromated flour if you do much baking, especially if I can't detect any real taste or texture differences between the two.

A very important quality of flour which is often not named on bags is the percentage of protein, especially gluten. Breads need relatively high gluten content. Cookies want a lower gluten content. All-Purpose flour has a lower gluten content suitable for cookies, but the precise amount varies by manufacturer and is bound to affect the way cookies turn out. Cake flour has the lowest gluten content of all. I'll need to find out from the manufacturers what the gluten content of these flours is, and may as well check on the bromation while I'm at it. I'll post the specifics of these characteristics of the flours at a later date.

One of the things I'm curious about is the moisture content of flour. There are many recipes that give a guideline for the amount of flour, but direct you to add flour until the dough reaches a certain consistency. I saw a cookie recipe claim that the amount of flour needed couldn't be given precisely because the amount of moisture in the flour varies at different times of the year. Flour comes in paper bags which are of course permeable to the air and its humidity, and so the statement in the cookbook sounds reasonable. It also sounds like an impediment to reproducibility of cooking results! I think I'll check on just how much the moisture does vary over this coming year. I did an initial measurement of moisture content last night.

To measure the moisture content, I sifted 40g of flour, or about one-third cup, onto my lightest cookie tray, and spread it fairly evenly across the tray. I preheated the oven to 250 F. and placed the tray with the flour in the oven. After ten minutes I removed it from the oven and weighed it. Then I put it back in the oven for another five minutes and weighed it again, and did that repeatedly until two successive measurements gave the same weight. I did that for each flour I have. The idea is that the heat drives the water out of the flour. The weight will decrease after it is heated for a while, and the difference in weight before and after heating should be the weight of the moisture it contained. I also confirmed the obvious, that heating the baking sheet I was using in the oven all by itself didn't cause measurable loss of weight, so that any weight difference occurring in the flour on the pan was definitely due to the flour losing weight, and not the pan.

Of course I don't know for a fact that the only thing that was changing while the flour was in the oven was moisture being driven off from the flour. I was essentially baking the dry flour at 250 degrees, and its smell did change during that time, so I'm sure there were some other things going on besides water being driven out. However, the color didn't change as far as I could tell. I don't think anything else that might have changed the weight much was going on. If combustion took place, it would cause loss of weight as carbon dioxide and water, but there was no color change or smell of burned flour or blackness, plus see the point in the next paragraph. I'm going to call it a reasonable measurement of water loss unless somebody who knows better tells me it is not actually measuring water loss.

In every case, the weight after ten minutes at 250 was significantly less than when I put the flour in the oven. The fact that the weight stabilized between ten and fifteen minutes (for two samples, between fifteen and twenty minutes) indicates that whatever was happening to change the weight was occurring in the first ten minutes, and then stopped. This is consistent with the "happening" being loss of moisture due to heating, rather than something like slow combustion that would continue with loss of weight until all the flour was consumed.

I was surprised at the consistency of the weight loss. In every case, the weight of the flour was reduced from 40 grams to 35 or 36 grams. My scale isn't overly precise, so I can't calculate the moisture content especially accurately, but taking the inaccuracy of the balance into account, these were possible extreme values:

(41-37) / 41 x 100% = 9.8% water
(41-35) / 41 x 100% = 14.6% water
(40-37) / 40 x 100% = 7.5% water
(40-35) / 40 x 100% = 12.5% water

So the moisture content for each of the flours was somewhere between 7.5% and 14.6% by weight, with the exact values unable to be known with greater accuracy using this scale unless I want to take the time to perform the drying procedure a whole pile of times with each flour and apply some statistical analysis to it, and I don't think it's worth it.

Let's suppose the moisture content is 11% by weight, about halfway between the two end values. How much water is that? Well, 40 grams is about a third-cup of unsifted (actually, "presifted") flour. 11% of 40 grams is 4.4 grams. Water weighs one gram per milliliter, so that's 4.4 milliliters of water in one-third cup of flour. Three cups of flour would hold 4.4 ml x 9 = 39.6 ml of water. The conversion for milliliters to cups is:

225 ml = 1 cup

and so 39.6 milliliters of water is 0.176 cups. In other words, three cups of any of these flours contains around three sixteenths cup of water, almost a quarter cup, but perhaps somewhat more or less because the weight measurements are inaccurate.

This is perhaps more water than I would have thought, but the numbers don't lie. Unless I miscalculated somewhere! Let me know if you see anything wrong.

Now, the amount of water I measured last night isn't really all that important. If there's moisture in the flour, there's moisture there, and that's what we cook with. What would be more important would be if the amount of moisture varies greatly depending on the time of year, as that one recipe author said. A recipe with 3 cups of flour I'd bake today would include about a fifth cup of water built in with the flour. If the amount of moisture in flour doubles at another time of the year, would a 3 cup recipe later include bordering on a half-cup of water? That would be quite a variation - and that much water in a recipe would surely affect the results. That much moisture difference in a cookie recipe would indeed make a big difference. I made these measurements in the winter, when humidity in Wisconsin is low. How much water would the flour contain in summer, when the humidity is high?

I plan on repeating this experiment this summer to see if the results are different - if I remember :-) - but here's an interesting thought. Does the flour from different manufacturers go into the bags at different times, that is, different months, or pretty much all at the same time? And is the humidity all over the country where the flour mills are about the same when flour is bagged, or is humidity different in different states, so some millers bag flour with higher moisture content than others? Let me explain what I'm getting at. I weighed all those bags of flour, and if you ignore the slightly light Robin Hood flour, their weights ranged from 2.264 to 2.282 kilograms. That's only a variance of plus or minus 9 grams from a midpoint of 2.273 kilograms, in other words a weight variance of only 0.4%. It seems reasonable to think that the moisture content of the flour would change over time to reflect changing amounts of humidity in the air to which it is exposed. It's also reasonable to think that the humidity in the air at the time the flour was packaged varied a fair amount from bag to bag, since these were all different manufacturers, many in different states who packaged at different times. We are thinking that, as the moisture content of the flour changes, the weight content would also change. Surely each manufacturer originally packaged 5 pounds in each 5-pound bag. All the bags wound up here in Madison, Wisconsin, which has a specific, uniform humidity in the flour aisle of the store. They would all gain or lose moisture and thus weight while they sat on the shelf according to the humidity here in Madison compared to the humidity they were in when they were packaged - but they didn't! They all still weigh the same, with only a 0.4% weight variance between bags! And the midpoint of the range is 5.05 pounds, which includes the weight of the paper in the bags, so the weight is pretty much right on at 5 pounds!

So logic is telling me that the moisture content actually varies very little with the humidity, and that the moisture content is relatively strongly locked into the grains of flour. It does not take up a lot of water from the air or give water up in a dry atmosphere. It's not surprising that 250 degrees would drive it off - that's what I would expect to happen. That doesn't mean it originally got into the flour by being absorbed from the atmosphere after the flour was ground.

This is just logic, not true fact, but We now have a hypothesis to work with - we are proposing that water is not given up and absorbed in significant quantity by flour depending on the humidity. One more experiment will tell us something important. I have taken 36 grams of flour which I dried in the oven and placed it in a cup in an air-tight container, with a half-inch of water on the floor of the container. There will be plenty of water in the air inside that container for the dry flour to soak up, if it's going to take it up. In a week I'll Weigh the flour in the container, and we'll see if it increases in weight and by how much. Any change in weight should only be due to taking material in from the air inside the container - which material would pretty much have to be water. Originally that flour weighed 40 grams. Will it go back from 36 grams to 40 and over? Or will it still weigh 36 grams despite the humidity within the container? We shall see. Stay tuned.

I'm not going to say any more about this until the adsorption experiment is completed. However, something is fishy about the weights of those bags of flour being so consistently close to five pounds, when supposedly the moisture content of flour changes depending on the time of the year.