## Sourdough Rise Time Table

I've had a number of discussions with TFL participants recently about sourdough rise times versus temperature and inoculation. Temperature has a big effect on sourdough rise times, and sometimes a starter appears unhealthy, when it is really just rising more slowly because of low temperatures in the kitchen during winter. Also, recipes that used to work seem to fail during the winter, but the colder temperatures may be the cause. To adjust for cold winter kitchen temperatures, either the temperature must be managed actively (oven with pilot light or electric light, coolers with a bowl of warm water in them, and so on), or the percentage of fermented flour must be adjusted in the recipe, or much more time must be allowed for the bulk fermentation and proofing.

I constructed a ** table** that provides (in hours) the doubling time, bulk fermentation time, proofing time, and total mix-to-bake time for various temperatures and percentages of fermented flour. The table has two sections, one for no salt meant for unsalted levains, and one for 2% salt meant for doughs or salted levains.

Inoculation, as used in the table, is the percentage of fermented flour contributed by a levain or storage starter to the total flour in a levain or dough. For example, if 50g of storage starter at 100% hydration is contributed to 225g of flour and 175g of water to create a levain, then the total flour is 250g (25g+225g) and the percentage of fermented flour is 10% (25g out of 250g total flour). Similarly, if a dough containing 1Kg of total flour is made by contributing the levain just mentioned to 750g of flour and 550g of water and 20g of salt, then the inoculation or percentage of fermented flour is 25%, or 250g out of a total flour of 1Kg.

The table is made to match up to rise times for whole wheat, high extraction, or generally high ash content flours I tend to use in my sourdough hearth breads. For pure white flour doughs and levains, the times tend to be about 20% longer, i.e. white flour rises a little more slowly.

Your starter may well be faster or slower than mine. If you build a test levain using a representative entry in the table, such as 10% at 75F, you can see how your starter compares to these table entries and then adjust your rise times and proof times up or down by the same percentage. For example, if you starter doubles in 80% of the time indicated in the table, then it makes sense to use 80% of the time in the table for other temperatures and inoculations also.

You can see from the table that the rise times vary over a huge range depending on temperature. Also, inoculations need to be changed drastically for long overnight rises, depending on temperature.

The strategy for maintaining a starter should also change dramatically if the temperature is 65F instead of close to 80F in the kitchen from winter to summer. For example, a 25% inoculation at 65F results in a 10 hour mix-to-bake time, which is a couple of hours before a levain would peak and begin to collapse, but at 80F an inoculation of only 0.5% results in a 10 hour mix-to-bake time. I've used this model at wide ranges of temperature and had reasonable results. The interesting thing to notice is that a 20g:30g:30g feeding at 65F peaks in around 12 hours but a 1g:100g:100g feeding at 80F peaks in around 12 hours, too. Or, if you look at the mix-to-bake time at 65F for a 10g:45g:45g feeding (10% inoculation), it's 12.5 hours, so if you feed that way at 65F the starter won't be getting to its peak and may be overfed if the feeding is repeated every 12 hours, while the same feeding at 80F will peak in less than 8 hours, so a 12 hour schedule will work well at that temperature.

This is simplified from my rise time models, so it doesn't include some additional adjustments for the dough consistency I make in my spreadsheets. Of course, this is a very rough approximation. All kinds of complications may cause these numbers to be different from actual results. So, it's just a guideline and something to think about, and it's biggest use may be as a learning tool or to just get in the general ballpark for rise times. For example, if your temperatures are very different from the ones the author assumed in the recipe, or if you just don't have an idea where to start with rise times for some recipe your trying, maybe the table will help.

Apologies in advance, if it turns out there is a bug in the table somewhere, but at least some of the numbers made sense after browsing through the table.

## Comments

Bill,

Thank you so much for re-posting this table; I remember you posting it before, but it went right over my head, so I promptly forgot about it!

I'm still trying to revive my sluggish pet, but it's coming along so I think I'll be able to run a test and compare it to your table soon.

Thank you so much for all your help on these things! You are incredibly generous with your time and knowledge to those of us flailing around in the flour and water...

edh

Many thanks for this, Bill!

I will stare at the table. (I'm afraid I am on overload right now and seem incapable of comprehending the numbers. So much for having been an honours student in math and science....)

-Elizabeth

P.S. And I thought that breadmaking WASN'T rocket science. ;-)

edh,Elizabeth,

I hope the table may help, even if just to put rise times in perspective vs. temperature and to understand how rise time varies with the percentage of fermented flour initially contributed to a dough or levain. I never posted this table before, but I posted a spreadsheet with some of the same models in it. I think the table is nice because anyone can use it without downloading a model and figuring out my inputs. Also, you can make your own adjustments to what it suggests, once you know how your starter and flours vary relative to the table results.

The table has groups of 4 numbers for each temperature (you can see temperature on the left side) and inoculation (the percentage of fermented flour across the top row). There are two tables side by side. The one on the left is with no salt, the one on the right with 2% salt.

The four numbers are:

In order to not overproof, the total mix-to-bake time should be from the moment you mix the fermented flour with the rest of the flour until you put it in the oven, including all kneading, shaping, scoring time. This is because the fermentation is going on during all that time, and you don't want it to go on any longer than the right amount of time.

Remember that white flours are generally slower to ferment than whole wheat or partially whole wheat flours. Wetter doughs will ferment a little faster than firmer ones. Everyone's starter is a little different. So, these are rough guidelines for a typical starter (mine) and a typical dough - not too wet, not too firm.

Bill

Hi Bill,

Could you take a look at your sourdough timing table. It doesn't seem to load for me.

Thanks,

Eric

Bill,

Quick question concerning table value - at a given temp as the inoculation % increase the mix to bake time decreases and the bulk fermentation also decreases - this is intuitive and makes sense to me. I am having a problem understanding why as the inoculation % increases the "proof time" also increases - it seems it should be decreasing due to the higher % of inoculant. I must be missing something - can you help explain this to me? This is a great aid - one I have been thinking about and looking for for a while - thank you for sharing your work.

Ben

Is the actual "live" Excel (I presume) file available for download? It would be more compact and manageable than the five images in the posts. Either way it is a work of art. Thank you!

Greg

I am trying to work out a schedule that will allow me to delay a couple of sourdoughs. I greatly appreciate the work you have put into the table and am very interested in what happens at fridge temperature. Will get back to you with results. Thank you for the guidelines!

Every once in a while, I encounter something so helpful on The Fresh Loaf that I want to bounce right out of my chair.

Such is this table of data.

(I couldn't find the link to the table at first, it's here: http://www.wraithnj.com/breadpics/rise_time_table/bread_model_bwraith.htm)Two Three questions.

1. If the final dough has the following composition:

What is 4.14 hours = ??(I know it says to use this value to compare against my starter and adjust accordingly, but I'm not sure I fully understand what that means). If my starter is, for example, 10% faster, or has a doubling time of 4.554 hours, would the following be correct:2. Why does the spreadsheet stop at 25% prefermented flour? I regularly make breads with 40% prefermented flour. I'm worried that, because it stops at 25%, I'm completely misinterpreting the purpose of the data (or at least the part that says "

Inoculation is the percentage of fermented flour to total flour in the dough or preferment when it is first mixed.") My example above (250 g prefermented flour of 1000 g), then I should be looking at the 25% column, yes?(It's the term innoculation that's confusing me, as I use that term for adding seed culture to flour and water to build up a starter. I innoculate some flour and water with a seed from my mother starter. When I add the preferment to the final or intermediate dough, I don't call it innoculation, although I guess I should? Is that what you mean my innoculation here, just adding preferment to final or intermediate dough?)3. Is the spreadsheet with cell formulae available for download, so I can see that math? If so, maybe I can extend the spreadsheet for myself to 50% prefermented. If not, how are the data derived?

Again, much thanks. If I'm understanding the data correctly (and the data are accurate), if could take a lot of trial and error out of my sourdough time estimates.

Thanks,

Thomas

Well, I know it has been a very long time, so you may no longer care. I happened to log back in to check something and saw this. I guess I missed the original notification somehow. As to question 1, yes that is my model's guess at the time it will take for the dough to double in volume. I think the any of the times would be proportional for a dough with a different doubling time, just as you describe.

As to question 2, my model has trouble predicting situations where the percentage of fermented to total flour is higher than 25%, since it is really just a guess at the growth rate of the culture based on a lot of experimentation I did at one point. I essentially codified the growth rates as a function of temperature, salt, hydration, and so on, but the average growth rate slows down dramatically as the culture becomes saturated. My model doesn't really handle that, as I did no experiments and did not fit any data for very high fermented flour percentages. I have used my model for a variety of hearth breads from ordinary to focaccias and pizzas to bagels with good predictions. In particular, I find the model extremely helpful to predict long fermentation at cool temperatures, when the accumulation of growth can vary by hours as a function of temperature and time in the refrigerator or incubator.

If I were to start over knowing what I know now, I would start with a model that integrates the growth rates and pays no attention to "doubling of volume" but instead tries to predict the accumulation of acid over time. I have read some papers on the subject and it wouldn't be that difficult for someone with some time to make a model with the same basic structure as mine for the inputs and outputs, but with an underlying concept of accumulating a varying growth rate, as conditions vary with that growth.

3. Unfortunately, I don't think my model would help that much since it is really based on the relatively constant growth that would occur during the phase when the dough is at lower concentrations of organisms. However, maybe you could extend the tables, if you did your own experiments. I would think the above suggested approach integrating the growth rate would be a much more flexible model that would cover a much wider range accurately and more intuitive to build, as well.

I am posting a link to a shared Google Spreadsheet. i don't know if it will work, as there are functions that would need transfer along with it, but I now only use my models in Google Spreadsheets. Hopefully, you could copy this spreadsheet and use as you like. There are many flaws with the approach in retrospect, yet I have used it, as I said, for many different dough recipes.

Sourdough Bread Spreadsheet by bwraith

<help>

I was updating my server, but I think the links are there again, although I am only a few minutes past having rebuilt it.

Hi Bill and thanks, I think we got some kind of esp connection... that was fast! Thank you again, Mini

It's been a long time since I've seen anything from you and it's a pleasure to know that you are still above ground. Hope to see more from you.

Paul

I've seen here and making me confident that my issues with prolonged bulk fermentation and flat fermentation in the fridge are really temperature dependent and not a fault of my starter. Incredible. Your table is incredible. Thanks for linking to it.

Thank you so much for this.

Hi hreik,

I'm glad you found this useful. I agree that understanding how big the variations of culture activity are with changes in temperature can resolve so many "mysteries" regarding rise times of dough or maturation of culture. I would add that when you are experimenting with "long rise times at cool temperatures", a closely related issue is the size of the dough and the insulation of the container being used for the dough during the rising periods. For example, if you have a very large dough, it may take many hours in the refrigerator for the dough to come down around 40F, and certainly the outside of the dough will be cooler than the inside. So, it makes sense to do some folding now and then, to even out the temperature in the dough as it cools down. It also makes sense to keep a chart of the temperature change, so you can measure a "heat transfer rate". In rough fashion, you should see that the temperature drop can be characterized as something like: "percentage of temperature difference per hour", for a given size dough and a given container (both the shape and the material of the container could have an effect on this percentage change in temperature difference per hour). For example, the dough starts at 70F, and the refrig is at 40F. This is 30 degrees difference. After one hour, you fold the dough (to get the temperature consistent and averaged throughout the dough, but measure the temperature in a few spots to make sure you have a uniform temperature, or at least take an average temperature), and let's say you measure the average temperature to be 60F. This means the temperature dropped 33% of the difference (10 degrees drop out of the 30 degree difference between the starting dough temperature and the final dough temperature after one hour). So, you should expect that after another hour, the dough will drop by 33% of 20F (60F dough minus 40F refrig), or about 7F more degrees, to a temperature of about 53F. Then in another hour, you might expect a drop of 33% of 13F (53 dough minus 40 refrig), or roughly 4F degrees down to about 49F, and so on. If you have a very big dough, then it will take longer, and the percentage drop per hour will be much less. If you have a really good insulating plastic container, the drop will be slow compared to a metal container or other material that lets the heat out of the dough much faster. Of course, it makes sense that a dough that is spread out like a focaccia will very rapidly drop in temperature, whereas a big round dough of the same volume as the focaccia may drop in temperature much more slowly. When I started doing long, cool, slow fermentations (pretty much my standard/favorite approach these days), I modified my models to use a "heat transfer rate" as described, and then add up the growth rate for temperature as predicted by the transfer rate, depending on whether the dough is out on the counter or in the refrig. It is difficult to do this without a spreadsheet, but you can get a reasonable guess by imagining what the average temperature is for each hour, and then adding up the percentages of the "doubling time" that each hour represents. You want those fractions to add up to 100%. For example, if you have a dough that rises at 70F for 2 hours, then at 60F for 2 hours, then at 50F for 2 hours, then goes back up to 60F for 2 hours and finishes at 70F for 1 hour, you would find what percentage of rise time 2 hours is for a 1% inoculation at 70F, and what that percentage is for each succeeding temperature and time. Then, add those percentages up, and they should come out to 100% as the doubling time for the dough. This is an extremely rough, sort of common sense way of looking at it, but the idea is that rise times are logarithmic with growth of culture, so you should be able to add up the percentages of time at different temperatures and get the same doubling of the culture.

As a point of comparison, I typically make doughs that have 6Kg of flour in them, and weigh something like 10Kg total. If I put them in a plastic food container that is rectangular, the insulation is pretty good. If I put it in the refrigerator about 45 minutes after the fermentation starts (when the starter is mixed with the some of the main dough flour - takes a while to mix all the flour and water and ingredients, run in mixer a couple of times, fold into a dough, drop in container), it will spend probably 6 hours getting down to something like 57F, after which I put it in my "incubator" that will maintain it at around 57F (the incubator fuse blows if I try to start with the dough in the incubator, so I put it in the fridge first to get the temperature down). Then, I let it sit at around 60F in the incubator for 16 hours or so. Then I shape loaves. They rise from around 60F up to around 68F in another 6 hours and then I bake them. The percentage change in temperature while in bulk fermentation (one big dough in a plastic countainer) is on about 14% of the temp difference per hour. But, when in separate loaves (shaped like torpedos now and in metal pans in a Ziploc Big Bag), then rise at more like 30% of the temperature difference per hour (first they are in a refrigerator at 40F for 6 hours, then in an incubator at 60F for 16 hours, then out on a table at about 72F for 6 more hours). The dough has risen about 65% when I shape them after the incubator stage. Then the loaves undergo the final rise for another 6 hours at somewhere between 60F and 70F. This turns out to work nicely for me pretty much every time, but if I changed the container, container shape, made a different amount of dough, changed the dough shape, changed the time or temperature of the various stages (have to do it sometime, if I have to go out for dinner, let's say, right when I was supposed to move the dough from the fridge to the incubator), well then I'll have to be very careful to completely figure out all those effects. Some of those effects are pretty hard to predict, although I've gotten pretty good at guessing by some common sense what changes in size and shape might tend to do to the heat transfer rate. OK, well I say all this because even with the insight of how much temperature affects things, then you have to still realize that this big factor (the temperature) in the dough is highly a function of all kinds of things over time, if you start playing with refrigerators and containers and shapes.

your previous post and the table, your response is again, amazing and incredible. Thanks a ton. Printing it out.

hester

Hi, bwraith. I admire the work you put into this table. I've been building up a spreadsheet of my own, and I'm at the point where I should figure out the adjustment in time required as I vary the percentage of starter. I see that you have that figured out, but I was hoping to work that formula into my own calculations. Can you tell me how that works, or point me to a source?

Hi jcope,

I make the simplistic assumption that there is a logarithmic growth rate. So, if N0 is the initial percentage, and N1 is the final percentage of organisms, where 100% is some "ripe but not too ripe" concentration of organisms (like the concentration when dough has just doubled in volume), then N1/N0=exp(r(temp, other variables)*t), in other words there is some exponential rate of growth as a function of temperature (and maybe salt concentration or other variables, in units of something like %/hour), so the ratio of starting to ending concentration is just "e" raised to the growth rate times the amount of time (all in sensible units). So, if you have an experimentally determined result that at some temperature of interest the growth rate is 10% per hour, and you think of 100% as when a dough has just doubled, then if you make a dough with 10% of that just doubled dough and remaining just flour and water, then you could calculate how long it would take to get back to just doubled by solving the equation 10 = exp(.1*t) for the t, the number of hours it would take to grow back to 100% from 10% (factor of 10 growth). It gets very murky figuring out exactly what 100% is, and how much more the concentration rises (or falls after a while) in a culture that matures beyond the doubling point, which determines what the ratio of N1 to N0 really is. However, in a normal hearth bread situation, the bulk of the time is spent in the "linear exponential growth" phase, and the effects at the beginning and end can probably be built into some adjustment you make to your model from experience. As I've said in the past, I might have tried to do something fancier with a growth curve that tries to quantify more precisely the conditions of growth over time and build that into some kind of differential equation, but I've just used simple linear assumptions with good success, so I never got around to doing it. I hope this helps.

Thanks for taking the time to respond, bwraith. I thought you would just tell me your 'r', which I could use as an approximation of my own, but I guess not. But on the other hand reading your ideas inspired me to figure it out myself. A couple things occurred to me: 1) I have a temperature-controlled environment (an 8 bottle wine cooler I use for bulk fermentation), and 2) every day I throw away some starter that I could be using for experiments like you describe.

I think each day I'll mix up small fixed sample sizes of dough at a fixed hydration percentage and measure its behavior with varying amounts of starter percentage. I can even do the autolyze step. I don't think it's necessary to measure time to double, in case they don't. Also time to peak will be inaccurate, because it's hard to find the exact point when that occurs. Time to reach any fixed point short of the peak will probably be ok. Also I can use my old iPhone to take time lapses that will allow me to accurately measure when the goal is reached.

Then I can graph it, maybe fit a curve, and use the results to fine tune my ferment and proof times to fit my schedule and to get the best flavor. I could also publish the findings on this forum for others to use, and then live the rest of my life on the royalties!

Thanks, bwraith.

Hi jcope, Sorry, I thought your question was more about how to calculate and what my table is doing, rather than about the actual rates. I guess you could easily calculate the rates in the table, if you just want a quick idea. For example I have a doubling time of 4.98 hours for a 10% inoculation with no salt in the dough at a temperature of 75F in the table (if it's the same table). So, you could solve for the growth rate as log-base-e of 10 divided by 4.98 hours, which comes out to 0.46, which could be interpreted as 46%/hour growth rate with no salt at 75F.

Your description of how you might go about doing this yourself is pretty much exactly what I did. I have an incubator that controls the temperature fairly precisely that I also use for bulk fermentations. Years ago, I did hundreds of experiments at different temperatures, hydrations, salt concentrations, inoculations, and so on, and calibrated my models to fit the experiments, after which I built a spreadsheet that I use each time I make bread to design the recipe and track the temperatures and growth as the dough rises.

You're right to question the accuracy of "doubling". I used graduated cylinders to help improve the precision and consistency. I also did experiments with a range of different hydrations, as the texture of the dough affects how fast it rises, somewhat independently of the culture growth rate. However, I found that with white bread flour and supple dough consistency (not sloppy wet, and not stiffly dry), that I could use the concept of doubling the volume as a good proxy for the growth rate. When you start with high inoculations, my approach has some problems, since you are getting into an area where the dough may not double before it collapses, and the growth rate slows down and is not that well correlated with how long it takes for the volume to change. That's why one of these days, if I ever have time, I would start over with a dynamic model and use pH or acid concentrations to track the fermentation instead of rise times. However, at the time, it was a quick practical way to get things working, and I have used my spreadsheet for many years for all kinds of recipes, even managing to tweak it (close enough) for situations that are not as well suited, like very stiff bagel doughs or very sloppy focaccia doughs.

I would encourage you to do some of these experiments and write everything down, make models and spreadsheets of your own. The process of running a variety of dough experiments will result in a really great feel for things, and building models results in really thinking out and organizing in your mind and on paper what is really going on (and there is truly a lot of complicated stuff all happening at once).

Meanwhile, I have moved all my spreadsheets to Google Sheets. A Google account is free and allows one to use Google Sheets. Here is a link to my latest spreadsheet. Most of the calculations and parameters to the model are in the supporting scripts. If you understand some programming, it's pretty easy to see what the calculations are by looking at the various functions in the script file. Some of the experiments are in the "model" sheet. The "bread calc" sheet is for specifying a recipe. It is possible to easily represent many different types of prototypical hearth bread recipes with the formulas as they are, and one could easily create other ones by just changing the formulas. Of course, it might take some careful examination of the formulas to understand and use the spreadsheet. I admit that this spreadsheet is not designed for "ease of use", lol. However, I have tried to insert useful "notes" that may help you figure out the formulas.

The "Summary Page" is where the actual timing is calculated. I have a section at the bottom that is for situations where the dough is large enough that it takes significant amounts of time for the temperatures to change, if you put the dough in a refrigerator or incubator. You can play with the "heat transfer rates" and the temperatures and times in the various stages to try to match up with the characteristics of your containers and shape of the dough.

The spreadsheet needs to be copied into your own account to be useful. I have set it to "view with link", which allows you to view but not change anything if you visit the version in the link. However, if you make a copy of it and use the copy in your own Google account, you can do whatever you want with it, including modifying it to suit your needs.

The link:

Sourdough Spreadsheet by bwraith v0.2

I realize that the spreadsheet itself may be hard to make use of, but maybe it will help in some way. I use it every time I make bread.

Here are links to two papers that were very useful in thinking about modeling sourdough culture growth.

Effects of process parameters on growth and metabolism of Lactobacillus sanfranciscensis and Candida humilis during rye sourdough fermentationModeling of Growth of Lactobacillus sanfranciscensis and Candida milleri in Response to Process Parameters of Sourdough Fermentation.Regards, Bill