Retarding Dough For Flavor Enhancement and Process Control

The information below is taken from an EXCELLENT discussion in the Bread Baker's Guild of America mailing list. It is hard to overstate how much I have learned from this group, and strongly encourage any serious baker, amateur or professional, to join and support the guild. Much of the information is from Abe Faber, the former owner of the Clear Flour Bread outside Boston. Abe is one of the keystones of the Bread Baker's Guild of America, and one of a handful of people to whom I happily defer.

The first part of this discussion is fairly technical, but at the end I'll bring it around to home practices for serious home bakers. However, I think the earlier information is very important to understanding how retarding dough works. I'll apologize in advance for this being long, dry and having no pictures. But, I think the information is very useful!

There are three essentially different kinds of possible process involving the use of refrigeration, which are all too often lumped together under the broad title, "retardation."

1 Retarding in bulk, otherwise known as a Delayed First Fermentation.
2 Slow Final Proof or rise of finished loaves.
3 Retarded Proofing process of finished loaves.

When retarding in bulk for 12-18 hours the dough will benefit from being down around 45-48 degrees. It can be divided and pre-shaped (if applicable) right out of the retarder. The dough should probably be rested a little longer than normal after pre-shaping as it's going to take a while for the dough to come up to temp, but it can be shaped and final proofed just the same as a regular dough. Three to four hours may elapse between removing from the retarder through divide and shape and proof and until out of the oven. Yeast can start at 1.2 percent for this method.

When retarding for a slow final proof or rise of finished loaves, the dough can be divided and pre-shaped just half an hour after mixing. Allow the dough to rest another half hour or so and then shape it. Then the loaves are set in a retarder at 50 degrees. Anytime between 12-15 hours later the loaves could be pulled out and baked directly. Yeast for this method start at 1 percent but the amount depends on the length of the intended long slow proof.

When using the retarded proofing process of final loaves process one does the same rapid divide, pre-shape and shape as described above but then stows the dough away at 38-40 degrees. Now the dough can sit in there anywhere from 12 to as much as 36 hours before pulling the loaves and finishing the proof for as many hours as needs to come up to room temp. Or by allowing a automatic proofer-retarder to come up to 75 over a few hours. Yeast for this is more near 2 percent.

Why do we use more yeast if we want to retard the dough? Because when we do wake it up we want it there to go! This is the method those of us who do not have neato multi temp units can use.

For both the methods where the bread is shaped up front you need a little stiffer mix and a little longer (more developed) mix, and definitely preferment to insure enough strength.

The difference between RETARDING at 38-40 as opposed to SLOWLY PROOFING at 50 - 55 is a pretty big difference. With one you are putting things close to dormant (yes I know its not completely dormant - don't go bleeding edge technical on me about this), intending a final proofing stage, and with the other you are just asking the bread to drive all the way to the oven nice and slow in the breakdown lane and directly off the exit for downtown Ovenville USA. Yeast amount, temp and timing become everything here.

With both techniques for making the long period occur in the final loaf you have issues possible with the outer skin drying out if not using some expensive high tech unit, as well as potential for skin blisters at the bake, as well as potential difficulty with transfer to the hearth.

With all three methods you want the initial mix temp to be more like 73 degrees rather than 75-78, so nothing takes off too fast.

When going for these long slow techniques the qualities of your flour become incredibly important to understand as well. Those qualities are magnified!

If you happen to be using a mill run that came in with a falling number of 240(very active) you will have much less success in controlling the long slow process than if you have a nice puttering along flour with a falling number closer to 280, which is desirable in this case. In addition you really need flour with the gluten quality to stand up to the long exposure to the degradation as a result of all that time the dough is subject to the action of the protease enzymes. As far as I understand, the protease keeps on gnawing on the gluten structure even at the lower temps.

Sometimes a little ascorbic acid is warranted for these methods. For retarding the final proof longer than 12-15 hours some kind of (hopefully natural) dough conditioner might be needed. But if you have a method that is working and then ceases to work don't forget to look at fluctuations in the flours enzyme activity or protein quality as a cause.

For a high hydration dough the bulk fermentation is probably the way to start. Pulling a high hydration dough off a linen its been on for 18 hours can be dicey, although I'm sure there's a way.

I fully understand that all sorts of variables are possible other than these three. But these three represent the different basic approaches you are going for in what you are expecting from the dough. The choice between delaying fermentation and then allowing it to happen at the end, as opposed to controlling it slowly over a long period. Anything else is some combo of the two. It would probably be helpful to first try and observe the basics before going for hybrid situations.

A full and meaningful discussion of trying to control any particular dough over a long period requires looking at the ENTIRE picture. This means the whole formula including percent of prefermented flour, total yeast percentage, profile of the flour, how big are the individual tubs of dough going into the proofer, etc.. Otherwise everyone is taking shots in the dark reacting to just one of many factors.

It's possible to get almost the exact same final loaf using many methods and timelines but all the factors of flour selection, final mix consistency, final dough temp, etc.. become more important to control.

Home bakers who want to experiment with retarding may find that their refrigerator doesn't have enough of a temperature range to be able to stay at the 48 - 50F range discussed above. The answer is to go to a home brewing store and get a refrigerator controller from them. This device lets you control a refrigerator or freezer very accurately.

The last question is how to decide which technique to use. Retarding, in the first and second senses, allows for wonderful flavor development and should be recommended to everyone.

Most of the emails I get from home bakers on the subject of retarding dough are related to process control. "I want to serve freshly baked bread, but with all the stuff I have to do on Thanksgiving, I just don't have time to make bead too! Can I just refrigerate the dough and bake it on Thanksgiving?" It is hard to explain that there is a lot involved if you're going to do it right. But, this is a process control issue, and the writers need to pick the option that most closely works with your needs.

Delayed first fermentation has the advantage that it allows you to minimize your refrigerated storage - you are looking for space for a tub with numerous loaves in it, not a space to put a large number of formed loaves which need space around them. You need to look at how many loaves you are talking about and how much space they will take up, either as a bulk fermentation or as separate loaves. Delayed first fermentation is also less likely to produce a blistered crust. Many Americans like the look of a blistered crust, but the French feel it is an indication that the baker didn't do a good job.

The last technique, the retarding of finished loaves is strictly a matter of production scheduling and does not improve the quality of the breads.

That's the end of Abe's carefully considered comments.  What follows are my less erudite comments based on our experience in the Colorado High Attitude Bakery.

In our bakery, we used the second technique, the slow final proof. We tracked our process very carefully and we would form our loaves, give them varying amounts of floor time, depending on how that bread responded, and then put them in the retarder. In the morning, we'd pull fully proofed loaves out of the retarder and bake them immediately. This worked very well for us. We handled the drying dough issue by covering our doughs with lightly oiled plastic wrap. Some people prefer oil cloth because it doesn't stick and is reusable. We never tried that, but we did switch to processing caps later.

Now let's talk in more detail about how to prepare a bread for a Slow Final Proof or rise of finished loaves. We start by making sure we are using a good, strong sound flour. Our dough is made slightly thicker than other doughs that are making similar products that won't be retarded. We increase the amount of riser slightly, whether that is sourdough or yeast. We gave our breads a full first rise as lots of flavor is developed in the first rise. After the first rise, we scaled, pre-shaped, rested and shaped our loaves. We then covered them, gave them a variable amount of floor time to start rising. Then they were wheeled into a walk-in refrigerator at 45 to 48F to rise slowly overnight. In the morning, the loaves were removed from the retarder, uncovered and baked.

If we found that the loaves weren't fully risen, we'd wheel them into the proofer. Over time, as we dialed in our processes, times, and temperatures, we all but stopped using the proofer for breads.

This process has a number of control points which we fine tuned to get the results we wanted.

The flour we used was a potential control point, but our flour worked well and our supplier had a limited number of flours available. The thickness of the dough was another control point. Thicker doughs won't rise as well, but will deliver more sourdough flavor. The dough temperature was another control point. Warmer dough will develop more quickly.

Floor time, or the amount of time the dough is sitting on a cart waiting to go into the retarder, is another key as it gives the riser more time to do it's work. If a bread is consistently under risen in the morning, we increased its floor time. It it was over risen, we decreased its floor time.

The next control points were the temperature of the retarder and how long the dough stayed in it. Since we were making a variety of breads, we couldn't adjust the retarder to match each dough's ideal temperature. So, we played with the amount of riser, dough development and floor time. Luckily, we had enough control in those elements.

Depending on your needs, you can give your doughs time after the retarder to rise. We found this wasn't optimum for us as the outside of the doughs tended to develop too soon and the inside too late. Most of our doughs were made into pan breads, which minimized the concerns with sticky dough.

The difficulty most home bakers face is that their refrigerator is far too cold to extend a rise. Their only options are to get another refrigerator or freezer to use with an external controller to allow them to use warmer temperatures. Home bakers can play with the amount of riser , the dough formulation, the floor time, and the amount of rise time after the dough is removed from the fridge. It usually isn't a good idea to raise the refrigerator temperature to 45 to 48, much less the 68F used by many bakeries in San Francisco to produce San Francisco Sourdough Bread as the rest of the stuff in your fridge might not like it.

If you have questions about retarding dough, please use the contact us page to ask us about your questions.