As I outlined in the previous article, one reason for decreasing extract efficiency when brewing higher-gravity brews is that some brewers boil the same volume of wort, regardless of the size of their grain bill. If you do so and use continuous sparging, the volume of sparge water you use decreases as your grain bill gets larger. As such, the larger the grain bill, the fewer sugars are rinsed from the grain bed after the first wort had been run off. The same can be true if you batch sparge and your pre-boil volume does not go up with bigger grain beds.

Equivalently, with an unchanging boil volume in a “straight” BIAB system — one in which the bag is hoisted from the mash tun/kettle and allowed to drip, but isn’t sparged — larger grain bills mean higher gravity wort. When the grains are removed, the wort retained in the grain bed is the same gravity as the free wort. The volume of wort retained by the grain bed is also larger with larger grain bills. As such, progressively larger grain bills equate to more sugar being removed from the kettle by the bagged grains. More wort is retained by the grains and that wort is of higher gravity.

As I mentioned, in order to preserve your extract efficiency, you need to ensure that you completely sparge your grain bed. However, for some brewers, this is either not possible or inconvenient. The next best thing would be a way to predict your extract efficiency for whatever original gravity you desire — and there is one way easy way to do this. The catch is that you need to have had taken good notes for your prior beers.

#### A Graphical Solution

If you collected exactly the same amount of pre-boil wort each time you brewed, you could plot points on a graph of extract efficiency vs. original gravity, then draw a curve to predict your extract efficiency at any gravity. The more data points you had, spread over a wider range of gravities, the more accurate your curve would be.

If you collected different amounts of wort, and perhaps brewed different volumes of beer over different batches, you can make one adjustment — plot extract efficiency vs. the ratio of grain weight to pre-boil wort volume. If you’re not fully sparging the grain bed, higher gravity beers will have a higher grain weight to pre-boil wort volume ratio.

In the graphs in this article, I show individual points plotted (below) and the same graph with a “best guess” line drawn in red (above). The red line will give you an estimate of your extract efficiency at that gravity or grain-to-pre-boil-wort ratio. When you’re planning your next brew, locate the expected original gravity (or grain-to-pre-boil-wort-volume) point on the x axis (the horizontal axis at the bottom). Draw a line straight up until it intersects your best fit curve — on my graph, the red line. Draw a line from that point straight left until it intercepts the y axis (the vertical axis on the left.) The point it crosses that line will be your estimated extract efficiency. You can use any units you want for the grain weight and pre-boil wort volume, as long as you are consistent. (Extract efficiency is always a percentage.)

If you use an advanced spreadsheet or statistical software, you can get an exact best fit line to fit your data. If you draw it on graph paper, as I did, you’ll just have to eyeball it. (You may not get the exact best curve, but with enough data points, you’ll be close enough for practical purposes.) Be aware that the line (or curve) may not be linear, especially at extremely high original gravities (or corresponding grain-to-pre-boil-wort ratios).

You will get good estimates of your expected extract efficiencies if you have multiple data points in the vicinity of your planned brew. If you are reading from a region that does not have many data points or is outside the range of prior data points, expect the estimate of extract efficiency to be less reliable.

—

*Related articles*

Declining Extract Efficiency at Higher Original Gravities

Chris, one variable that this does not account for is the expected yield of the grains in the mash. While most pale and pilsener malts have a measured extract yield (in the lab) of around 80%, crystal malts are lower in yield and roasted malts lower yet. So grain bills high in these grains may yield less extract. This could be exacerbated if the brewer fails to account for pH changes that a grain bill high in crystal and roasted malts will have.

I am interested in developing methods for predicting mash efficiency. I have developed one for batch sparging that serves me well. I will email you separately with a description ( its is a little complicated to explain here). I would be interesrted in your thoughts.

All the variables that affect extract efficiency are accounted for when the extract efficiency is calculated. This then becomes the y axis. The x axis is a measure of how fully sparged the grain bed is (either OG in the special case of an invariable boil volume or grain weight divided by pre-boil volume).

This method adds one more variable (how fully sparged the grain bed is) to the existing variables that are accounted for when the brewer calculates his extract efficiency. If each plot comes from the same brewer, brewing on the same system, it should yield a good estimate for how much the extract efficiency needs to be adjusted according to OG or “completeness of sparging.”

Differences in the percentage of specialty grains and roasted grains might play a small role here. If you wanted to account for that, you could always plot each point with a colored dot to reflect the beer’s color. (A stout would have a black dot, and amber ale would have an amber dot, a Pilsner would have a yellow dot, etc.) You could draw separate lines or make separate graphs for beers made with all base malts, stouts and porters, beers with roughly 10% crystal malts, etc. I don’t think that will make a big difference, though.

While you don’t mention it in this article, I find this thinking useful for parti-gyle brewing with batch sparging. I fill a mash tun with X pounds of grain and have roughly 31 * X gravity points (assuming 1 lbs of my grist ground to congress mash specifications would yield a gravity 1.031 — pretty standard). I know I want to get P percent of the sugar into my first batch, so I can use the efficiency curve to figure out how much water to mash and sparge with such that the wort in the mash tun has the gravity I want. I then drain the volume I want (assuming it is possible to get) and add the second batch sparge volume to get close to my max efficiency. With a 70qt cooler mash tun, I have the room to make this happen for a wide range of 10 gallon (total) parti-gyle batches.

For this thinking, I find it easier to consider efficiency on the y axis, as you do, but have quarts of water/pound of grist on the x-axis. I’ve recently switched equipment, so I don’t have the data yet, but I suspect it would be from 50% to 75% on y axis, over the range of 1 qt/lbs to 3 qt/lbs, without appreciable grain above that point.

Whats the lowest water to grist ratio you’d use? 1qt/lb? Obviously it would boost your sparge volume, but at what ratio is too low for brewing? Wondering about the compensation for larger volumes that way…

Kai Troester did some tests of different mash thickness at two different water to grain ratios and then sparged the to get the same total water usage for both (I think — his description could be a little clearer): http://braukaiser.com/wiki/index.php?title=Effects_of_mash_parameters_on_fermentability_and_efficiency_in_single_infusion_mashing

See the “mash thickness” sections. He claims that thinner mashes lead to more efficiency, but I think his data are inconclusive on this point.

Around 1 qt./lb. is the thickest mash I’d use. (Just in case there’s some confusion, the grain weight to pre-boil wort volume I mention in this article is not a measure of mash thickness, it’s a measure of how much wort you collected from how much grain. I probably don’t need to mention that, but there are two ratios involving grain and water so I want to be as clear as possible.)