Making my own tempeh

I update this post from time to time as I learn more from experience. Most recent update May 5, 2021 to add a link to an article on sources for starter culture. An article I wrote in Medium distills the practical essence of the method I use and contains sufficient information for successful tempeh production. Update: I distilled the basics of the method I use, including refinements based on what I have subsequently learned, into this post, which also includes a list of the various tempehs I’ve made with links to the posts where I describe what I did, including photos. At the present time, I am using a tempeh starter culture from TopCultures.

---

.
I decided to make my own tempeh because (a) tempeh is very nutritious, and (b) I enjoy making my own foods (examples: mayonnaise, ketchup (for both, see this post), Worcestershire sauce, and pepper sauce/hot sauce (you can find my posts on those by searching the blog). Also, my current diet is the whole-food plant-based diet set out in Part 2 of How Not to Die, and among other foods, it has three servings of beans (or lentils) a day, and tempeh is a good way to satisfy that requirement. (More — much more — at that link.)

For more information on tempeh, see this useful collection of summary of scientific articles. And for a summary of what I learned and the method I finally arrived at, see this post.

I have made the tempeh in a 9×13 pyrex baking dish, so that it forms a slab. I’ve also used large Ziploc bags which I perforate: roll or fold the bag into several layers, then use a small-hole punch (1/16″ holes) to punch a series of holes through the several layers along the strip. I now use large Ziploc “fresh-produce bags,” already perforated, and those work very well indeed. I lay the filled bags on a wire cooling rack to help with air flow. It’s not necessary to press the tempeh: the mycelium will bind the beans together into a solid slab.

I use the number of cups of uncooked beans/grain to define batch size, though of course the beans are cooked as the first step in making tempeh. At first I made a 2-cup batch (2 cups of uncooked beans), but then tried a 3-cup batch. With the 3-cup batch I discovered that the tempeh itself generated heat once the mold was established. However, 3-cup batches work fine if I remove the batch from the incubator after the fungus is well started (12-24 hours, when a haze of white mold appears on the surface). Once the fungus has taken hold, I remove the batch from the incubator and place it on a raised rack on the countertop and let it continue at room temperature. Letting it proceed at room temperature also cured the sporing (black or gray patches), so I do that regularly.

I like to cut off a small piece tempeh, about 3″ x 4″ and cook that like a hamburger patty: a hot skillet, add a little olive oil and then the “patty” and brown it on both sides. With the 3-cut batches, the “patty” is rather thick, so I slice it into two thinner patties, and often I will simply dice those and cook as a stir-fry.

I top my veggie bowl with the cooked patty (the way some people top a veggie dish with a fried egg or use the cooked diced tempeh in the bowl. Sometimes I dice the cooked patty and use it as croutons in a salad. Diced tempeh also works well in stews, stir-fries, and chilis. I’ve also recently started making tempeh breakfast sausage. (I tried adding minced mushrooms, but then the “sausage” didn’t stick together — better to serve mushrooms on the side.)

I had a series of failed batches in a new apartment, and the reason (I at first thought) was that the incubation temperature was too high: the new apartment has a small oven, and with the light on and door closed, the tempeh mix just got too hot. My daughter suggested that a proofing box would work well (YouTube offers several DIY versions), though modern ovens often have a proofing setting.

I decided to make my own homemade tempeh incubator. Version 1 — made by applying closed cell foam to a cardboard box — it worked well so far as the tempeh was concerned, but cardboard is incompatible with the humid environment that the fermentation produces, and ultimately I judged Version 1 a failure because the lid pieces collapsed.

Version 2 I made using rigid 1″ foam insulation board. I used the seedling heating map and digital thermostat I got for version 1, since those worked fine. This post provides step-by-step assembly instructions, includes links for materials, and includes lessons learned from experience. The incubator works great and was easy to make. It’s a relief to have a reliable incubator — tempeh is a great food.

Here are the ideal temperature requirements:

During the first 12 hours, the inoculated beans are just heating up and no cooling is required, assuming the temperature of the incubator is around 88 °F (32°C). After 12 hr, a small amount of water begins to collect on the inside of the bags. At 14 hours, fermentation of the Rhizopus spores has begun in earnest and active cooling of the air must take place.

[Active cooling  is needed for industrial-sized batches. In my small homemade batches, using 2 cups of beans, no cooling is required, but I do cut back my incubation temperature from 88ºF to 77ºF after the mold is established, about 12 to 24 hours after start the batch. When I made a 3-cup batch (cooking 2 cups black beans and, separately, 1 cup black rice, then combining), enough heat was generated that simply turning down the thermostat was not sufficient: if the lid of my (version 2) incubator was in place, the temperature inside kept climbing. At first I just put the lid on a slant so that the incubator was partly open, but ultimately I had to remove the lid altogether because the tempeh was generating so much heat. I finally took it out of the box altogether, and when the internal temperature reached 99ºF, but it still seemed to work well if I left it on a raised rack at room temperature. – LG]

If the inoculated beans reach a temperature above 92°F (33.3°C), conditions are no longer ideal for Rhizopus spores but are ripe for a different set of organisms such as those of the Bacillus group. Rhizopus can be severely damaged by heat of over 92°F ( 33.3°C).

[As a point of interest: I used my new digital thermostat to measure my oven temperature with door closed and light on. It was 94ºF, so in fact building my tempeh incubator was indeed a good call. And when I maintained a temperature of 88ºF after the tempeh had taken hold, the mold started to spore, which resulted in unsightly black areas, though the tempeh was still good to eat. So it’s important to drop the temperature after the mold is started. – LG]]

At temperatures above 96°F (35.6°C), Rhizopus is killed and the product will be discarded. If internal temperatures don’t reach 96°F (35.6°C), the tempeh can usually be saved. Heat-damaged tempeh may be taken out of the incubation room and kept out at room temperature (approximately 70° F or 21.1° C), and it will start to grow over the damaged areas with new white mycelium.

The ideal relative humidity for incubating tempeh is between 50% and 75%. Airflow should be kept in the 120-cfm range to prevent overdrying of the product, which can lead to premature sporulation.

However, even with better temperature control, a new batch failed, and I had to reconsider possible causes. I then realized that around the time I moved into the new apartment, I changed my method of cooking beans. Instead of soaking the beans in plain water, I soaked them in brine (1 teaspoon of salt per 1 cup of dried beans) and instead of cooking the beans in plain water I added 1/2 teaspoon of baking soda (bicarbonate of soda) to the water.

Those changes do result in quick cooking and tender beans, but the alkaline cooking water makes the beans somewhat alkaline and thus toxic to the mold that produces tempeh, which requires an acid environment —  the reason one adds 1 tablespoon of vinegar per cup of dried beans to the cooked beans after they have been dried (or adds 2 tablespoons of vinegar per cup of dried beans to the cooking water for the last 20 minutes or so of cooking).

I had thought that the vinegar would counteract the baking soda in which the beans had cooked, but it did not — and the result was a string of failures. The current batch — soybeans soaked and cooked in plain water, no additions — turned out well once more. (I will say that I had to cook the soybeans a lot longer without the baking soda, but of course the time is worth it to avoid tempeh failure. But when I cook beans to eat (not use as tempeh fodder), I shall continue to use baking soda.)

Drying and cooling the beans

I worked out a good method of drying and cooling the beans: When the beans are done, I drain them through a sieve and spread out on a clean dishtowel, then use a paper towel to dry the beans. I place the paper towel over the beans, press gently, and rock the beans back and forth to dry them. I also sometimes use a hair dryer on the beans, being careful not to blow them away. With the beans spread out into a thin layer (1 bean deep), they cool quickly. Even if they get cooled to room temperature, that’s not a problem since the incubator box’s steady 88ºF will warm them up. Put the dried, cooled beans into a bowl, and if you did not use vinegar in the cooking water, stir in vinegar (1 tablespoon per 1 cup of uncooked beans), add the culture, and stir to mix, then load into a Ziploc fresh-produce bag  and place it on the rack inside the incubator. Spread the beans to fill the bag when it’s lying flat: you want a thin layer to avoid heat buildup in the interior of the batch. Note: here are places you can get starter culture. A bulk amount is much cheaper per batch. I store my starter culture in the refrigerator.

Making the tempeh

For tempeh, I like the idea of using my own choice of beans/lentils and adding enhancements as I want (for example, my first batch: pinto beans with flax seeds and cumin seeds and minced jalapeño).

This video shows one approach, and the video content is also in a blog post at Veganlovlie. That post also provides some details not included in video, so I recommend reading it. I follow her advice and don’t bother dehulling beans (and that seems to apply mainly to soybeans, which I’ve not yet tried but will in time). As noted above, I have found that it’s easier to dry and cool the beans if you spread the cooked beans out onto a clean dishtowel. It doesn’t matter if they get too cool because they will warm up in the incubator or oven.

Here’s a brief summary of the first four batches:

Batch 1
Pinto bean, flax seed, cumin seed, and jalapeño: 4 sandwich size zip-lock bags plus beans in 2 open dishes. Successful except that jalapeño should have been added after cooking since the flavor was attenuated in the end product. But good, solid tempeh. 

Lesson learned: Add minced jalapeño after beans are cooked.

Batch 2
Peanuts and kamut with curry powder. A semi-success. Curry powder contains ground coriander, an antifungal, and that undermined the mold.

Lesson learned: Check ingredients to verify no antifungal properties. Many spices and some herbs are antifungal, perhaps why they were used in preserving foods.

Batch 3
Black bean and green lentils with chia seed. A failure. Chia seed, which I added right at the end of cooking the lentils, forms a glutinous, pulpy coating that holds water. The mold never had a chance.

Lesson earned: Follow the instructions. The instructions in the starter packet state:

Dry the [cooked] beans by patting with a clean towel or using a hair dryer on low heat. Beans must be dry to the touch before continuing.

I now dry the beans by pouring the contents of the pan through a sieve to drain off the hot water, shaking the sieve a few times, then spreading the beans out in a then layer on a clean dishtowel (often on two stacked clean dishtowels). I then rub/roll them gently with a clean paper towel, and let the heat of the beans evaporate excess moisture. I leave them on the dishtowel until they are cooled to about room temperature (since they’ll warm up in the incubator). draining them, returning them to the pot, and heating them, stirring more or less constantly, until excess water has evaporated. I bought an inexpensive hair dryer and use that after turning off the heat, stirring the beans with a spatula as I blow hot air over them. Then I let them cool in the pot.  (Once they’re cool, I pick them up in the towel and transfer them to a bowl. I then add vinegar (1 tablespoon per cup of dried beans), stir well to mix (using a soft silicone spatula) and then add the tempeh starter, a little at a time, stirring well after each addition to ensure the starter is evenly distributed. I then bag the beans in a large Ziploc Fresh Produce bag (already perfectly perforated) or them spread out in a Pyrex baking dish (though nowadays I no longer use a baking dish).

Batch 4
Black bean and green lentils, with no fancy stuff. I cooked 2 cups black beans and 1 cup green lentils, so 1/3 larger than the 2-cup batch of the instructions. I decided to go with all bags this time, bags well perforated with my small-hole punch, each bag ~11 ounces. I used no weight on the bags of tempeh: I just put them on a rack in a baking sheet (to ensure good elevation). They firmed up fine as the mold grew.

Batch was a success, though the mold was not so dense and binding as in the open dishes of batch 1. I think, in retrospect, that this was because I insufficiently dried the beans and lentils.

Lessons learned: The open-dish method seems to work better than the punctured-baggie method. Also: heat the beans after they have been drained to make sure they are dry. Residual moisture counteracts the mold noticeably (comparing this batch 4 to the next batch 5).

Batch 5
I used red kidney beans (because I have a lot on hand), and I cooked 3 cups dried beans instead of 2 cups. In the video above, she adds 4 tablespoons apple-cider vinegar toward the end of cooking (for 2 cups dried beans). I therefore used 6 tablespoons (for 3 cups dried beans), adding the vinegar to the water toward the end of cooking. I had also added vinegar toward the end of cooking in batch 1 (the best of the first four batches) but not batches 2 and 4 (batch 3 being an utter failure and discarded).

[Note: I now no longer add vinegar to the cooking water. I cook the beans, drain them, dry them (heat pot and then using a hair dryer), and then add the vinegar, 1 tablespoon for each cup of dried beans I used, and the tempeh starter.

Another note: I add 1 teaspoon per cup of dried beans to the soaking water, and after I drain that water and refill the pot with fresh water, I add 1/2 teaspoon baking soda (bicarbonate of soda) to the water — this makes the beans tender and shortens cooking time.]

After the beans were cooked, I did heat them in the pot to ensure that they were well dried. Then I added 3 tablespoons brown-rice vinegar as suggested in the starter instructions. (They suggest 2 tablespoons but that’s for 2 cups dried beans, so again I increased the amount by 50%.) I used the “open” method, spreading the beans on the bottom of a 9″x13″ Pyrex baking dish—they made a thick enough layer using the entire dish. I used the open-dish method because that was a good success in the first batch. This post shows the progressive development of batch 5. When the tempeh was done, it had structural integrity from internal cohesion and required a knife to cut it.

Lessons Learned: The beans on top and edges dried out a bit, though they will still work fine (particularly, I think, in a chili), but next time I will use a loose tented-foil cover, with some slits to admit air. The big Pyrex dish worked much better than ziplock baggies, probably because more of the mold has access to air. Another lesson learned: I must stir in the starter more thoroughly (see the 24-hour photo in the post linked above.)

Batch 6
For Batch 6, I used 2 cups dried red kidney beans and 1 cup whole kamut. I cooked them separately and then combined. I took care to insure that they were dry to the touch before I added the vinegar and culture. This time I used white vinegar, which did fine. Because I add vinegar at the end, I decided to skip adding vinegar during the last stage of cooking. Interestingly, the mold initially favored the beans over the wheat: photos at 24 hours, 48 hours, and 72 hours.

I also noted that the tempeh starter instructions say that after 12-24 hours, the tempeh can be moved to room temperature because at that point it’s generating its own heat. So after 48 hours I took it from the oven and left it on the countertop, loosely tented with aluminum foil in which I had cut ventilation slots. At 72 hours I called it done. In the photo linked above, you see black and grey areas, but those are fine, just the sign that the mold has begun to sporulate. The slab felt thicker—probably because the extra 24 hours gave time for more growth of the mold.

Lessons learned: Remove the tempeh from the oven-incubator after 24 hours. Continue with the large-dish method, but tent loosely with foil. There’s a fair amount of humidity once the mold gets going, so make sure the foil has plenty of ventilation slots. By using the foil, there was no “dried-out” parts.

Batch 7
Green lentils for batch 7, nothing fancy. I used 3 cups of uncooked lentils, simmered them until they were just tender (they will continue to cook for a while after they are drained since you don’t rinse them with cold water, which would stop the cooking), made sure they were dry (let hot water evaporate as they sat in a colander and also poured them onto paper towels), cooled them to 95ºF, poured them into the Pyrex 9×13 baking dish, and mixed them with 3 tablespoons brown-rice vinegar and a packet of starter. Photos: start, after 24 hours, after 48 hours, and I decided it was done at 48 hours.

I then loosely covered the dish with a pierced aluminum-foil tent. After 24 hours in the oven with the light on, I moved it to the countertop, where it did indeed generate plenty of heat (bottom of dish was quite warm to the touch and when I checked the internal temperature using a digital probe thermometer I found it was 106ºF after sitting out overnight). After 24 hours, it throws off a lot of moisture, so I periodically turned over the tent to let the condensed moisture evaporate and ultimately I cut an exhaust hole about the size of a quarter in the center of the tent, something I will include as part of the process from now on.

Lessons learned: Remove tempeh to countertop after 24 hours in oven incubator. Cut a large hole in the center of the foil tent in addition to the slotted perforations made with a paring knife. Indeed, the tent may not be needed at all. The beans ended up wet in the center with less of the tempeh mold there, so the venting seems inadequate. Next time I do lentils, I won’t use the cover after the first 24 hours.

Batch 8
Soybeans at last, the traditional tempeh bean — and no fancy stuff this time. I cooked 3 cups dried organic soybeans, didn’t bother with any de-hulling, and made sure they were reasonably dry. I did not add vinegar during cooking; after the cooked beans had be drained and dried, I mixed in 4 tablespoons white vinegar and one packet of the starter (which seemed puny compared to the amount of beans, but the mold will grow).  Photo: at start; and after 24 hours and after 48 hours and after 56 hours Note the use of a dishtowel as a cover after the batch’s been moved from oven to countertop.

It is very tasty — better, I think, than the other tempehs, which perhaps is why soybeans are generally the bean used. Certainly the next batch will be soybeans.

Lessons learned: Use a clean dishtowel to cover the tempeh. Leave it in the oven incubator (i.e., oven light turned on) for 36 hours and then try the countertop (with dishtowel cover in place). Use soybeans again — they’re really good.

Batch 9
Another soybean batch, and this might be the last batch I log because it is so uneventful. I feel at this point that I know what I’m doing. I took 4 cups of soybeans this time, and soaked, cooked, drained, dried, and cooled them to 95ºF. Although the packet of starter says it’s for 2 cups of beans, I figure that the mold would grow in any case. I poured 1/4 cup (4 tablespoons) of white vinegar over the beans, mixed well, and mixed in the starter. I did not add any vinegar during cooking; from now on, I’ll add vinegar only after beans are cooked, drained, dried, and cooled. The beans in the 9″x13″ dish seemed too deep, so I brought out an ancillary 9″x9″ pan and transferred some beans to that so that the layer was only about an inch deep in both. I used the dishtowel cover, and put them in the oven-with-light-on incubator. You can see the baby pictures (after 24 hours) in this post.

The first batch

I used this tempeh starter. It works fine if you follow the instructions and avoid antifungal ingredients. I am going to try starting a new batch by mixing in some of the mold from the old batch.

I’ll describe the first batch in detail; subsequent batches followed the same patter. I put two cups of pinto beans and 1/4 cup brown flax seed into a pot, covered them with water and let them soak overnight. I would have included also 3 tablespoons cumin seed but I used the last of my jar last night in making roasted carrots, so I had to buy more. I added the cumin seed before I cooked the beans, but soaking them with the beans would be better, I think.

I added the minced jalapeño when I started to cook the beans and simmered the pepper with the beans. That was a mistake: the jalapeño s not noticeable in the finished tempeh. I should have added the minced jalapeños after the beans were cooked and drained.

The Eldest offered some intriguing suggestions:

• French green lentils and white beans with bouquet garni and mustard seeds (see below);
• Indian kidney beans and black lentils with garam masala;
• Moroccan chickpeas with ras al hanout;
• Ethiopian red lentils with berbere.

I made a batch using raw peanuts (without the shell) as the legume (batch 2). That was tasty and fit with Thai food (which uses peanut sauces). I use raw peanuts and simmered them for a couple of hours and then combined them with cooked whole-grain kamut.  I used the  2 tablespoons brown-rice vinegar (called for by the starter instructions) and also 2 tablespoons curry powder [the bad idea: curry powder includes ground coriander, an antifungal that stymied the mold .

Having learned from that experience, I dropped the idea of using mustard seed: mustard oil is an antifungal, and mustard seed contains mustard oil. (At for the third batch I checked in advance.) I used chia seed along with the green lentils (1 cup dry lentils and 1 cup dry beans, cooked separately) and the fatal two tablespoons chia seed, which ruined the batch by making the mix wet.

Resuming the summary of my first-batch effort. Once the beans, flaxseed, cumin seed, and jalapeños had simmered until the beans were almost done, I added  4 tablespoons apple cider vinegar  (as called for in the Veganlovelie method) and continued simmering until beans were tender. I no longer do that. I follow the starter instructions and just add 2 tablespoons brown-rice vinegar (not sweetened) to the beans after they’ve cooked and dried.

Once the beans had cooled to 95ºF, I followed the instructions included with the packet of tempeh starter (which contained 3 foil envelopes of starter, enough for 3 batches — normally it contains 4 envelopes, and when I told them I was short one envelope they made it good), adding 2 tablespoons brown rice vinegar to the dried beans.Then I stirred in 1 packet of starter.

I used a bamboo cocktail pick to poke holes in two ziplock bags for that method, and I also used two glass storage containers for that method (see video above). I subsequently ordered a small-hole (1/16″) punch to make ventilating the plastic bags easier. That works quite well. Indeed, I think you could use a regular (1/8″) hole punch.

Best way to perforate bag: roll it up into a fairly tight roll, about 1″ in diameter, flatten it, and punch holes through all the layers at once. When you unroll the bag, it will be fully perforated.

Here is the “before” photo of the beans in the glass containers:

I put the beans in the oven with the oven light on,  the two bags pressed between two cutting boards and the two glass containers on top under a large dome.

Update after 19 hours

A friend indicated that this looked unappealing, but I know that she likes Roquefort and gorgonzola and other mold-ripened cheeses, and she also likes bread, made with yeast which (like mold) is a fungus. And I personally like mushrooms a lot, so I am comfortable with culinary fungus, and tempeh mold is just another example—and a rather interesting and tasty one.

Update at 29 hours

I am going to let it go tonight and take it out tomorrow morning. The tempeh in the ziplock bags has at this point not done so well for, I think, two reasons:

1. The holes were too small. I used a bamboo cocktail pick, thus the purchase of the small-hole punch mentioned above.
2. Mashing the bags between cutting boards more or less seals the holes so the mold does not get enough air.

I decided ultimately to forego the weighting and simply place the bagged tempeh on a  tight-grid wire rack resting on a baking sheet..

The tempeh in the glass storage container, though, did quite well and formed a solid brick.

38 hours & done

I’m calling this done. Here’s the tempeh in the containers:

And here it is removed from the container and turned over. It is indeed a solid block:

I tasted a slice. Quite good, and the cumin is noticeable though not (in that taste) the jalapeño. Perhaps crushed red pepper would be a better bet (provided it’s not a antifungal). Or add the minced jalapeño after the beans are cooked. Maybe I’ll also add a dash of Wright’s liquid smoke next time.

The tempeh in the bags I’m leaving in a little longer since those got a late start. for the last 24 hours  I took off the top cutting board and let them breathe.

I call it a success.

Tempeh in bags, 42 hours and done

These were quite solid because, I presumed, they were pressed during fermentation. Now I think it was not so much the pressing as the beans being well dried before adding the brown-rice vinegar and the tempeh starter. I repeated the bag method again, but with bigger holes.