The Ueda Koji Method
This guide is an adaption for the homebrew scale so you can experiment before bringing it into the kura.
一麹・ニ酒母・三造り
Ichi koji ・Ni shubo・San zukuri
“Sake is a beautiful thing, both for those who make it and for those who love it.”
– Morikuni Ueda
Background
Throughout Western Japan, many sake breweries have adopted a style of koji making that uses large wooden boxes or plastic tubs (タライ tarai) to grow koji in rather than wooden trays. This method allows them to produce high quality koji for ginjo and daiginjo brewing, in larger amounts than the tray (futa koji) method using less labor.
A version of this koji making style was refined by Dr. Morikuni Ueda alongside a group of sake brewers, and they formed a working group around this method in 2015. In 2020 the group published an article [1] that explains the basis for Ueda’s koji making and details the complete method. It goes well beyond the generic タライ麹 “Tarai koji”, promising reproducible results.
The purpose of this article is to transfer and apply the same concepts at the home brew scale. Tips and key differences have been developed by our own working group of home brewers, in order to effectively carry out the Ueda koji making method at a smaller scale than the 15kg tubs used in breweries.
This guide has been expanded to serve as a stand alone set of instructions. The original article is well worth reading for further context, and for the development story of the Ueda method. Please contact the Sake Makers Discord to request a translated copy.
While some of the techniques described may seem beyond the reach of a beginner, with diligence they are intended to help anyone reach a high level with their koji making.
As Dr. Ueda often said, “This technology should not be monopolized by a few people, but should be spread widely so that many breweries can make good sake.”
Goals and advantages
To improve the quality and consistency of koji making for all styles of home brewed sake.
While the Ueda method targets koji for ginjo and daiginjo brewing, the techniques can be applied to other types of koji made for sake. With minor adjustments a full range of desirable koji types can be produced, from so-haze through to tsuki-haze. Note that tsuki-haze koji is only used for specific brewing methods, and its production requires the use of highly polished sake rice varieties that may not be available to the homebrewer.
Fortunately the same techniques can yield great results when using other types of rice and polish rates that are accessible to all home brewers.
Advantages of the Ueda method:
Increased enzyme yield and better enzyme balance for sake making, with better haze penetration.
Less precise soaking absorption target due to a moisture adjustment period after steaming.
Better koji temperature progression, with a far lower chance to be above or below the target temperature range during peak growth.
The timing for maintenance work becomes routine and the work is easier to do.
More consistent results from batch to batch, with better control over the finishing moisture content.
Possibility to scale up the overall quantity of koji produced for a given size of koji chamber.
Less likely to end up with ‘hine-koji’, overdeveloped / old koji.
Equipment required
Cooling racks covered with clothes or mats to spread out the steamed rice in a 25mm / 1” layer.
Plastic or other non-absorbent tubs of a single size with lids or other covers. The tubs should be large enough to hold the desired koji amount with room to spare, eg. a 2 litre container can hold up to 1kg (dry weight) of rice. Tub height should be 80-100mm or 3-4”, avoiding tall and narrow containers.
Cotton or natural fiber towels or cloths for covering and insulating the tops of the tubs.
Remote probe thermometer or instant read thermometer for periodic temperature checks.
Digital weigh scale (5kg capacity) and calculator to record weight % amounts.
Small digital weigh scale for measuring the koji starter, 0.01g resolution.
Reliable humidity sensor or wet bulb thermometer to monitor chamber conditions.
Notepad or spreadsheet to track batch notes.
Detailed steps
These steps follow the same structure as section 3 of the original article. The procedures have been changed to make the process work better for smaller scale koji making, while applying the same control points that are central to the Ueda method.
Koji Chamber - Must be able to evenly heat up to 40℃ with moderate ventilation to keep the humidity below 50%. Many koji chamber heaters are not powerful enough to maintain these target conditions for drying the rice (step 3), in which case there are other options such as a low temperature oven. Whatever the heat source, try not to exceed 50℃ for a significant length of time since that will dry out the rice at the surface and can quickly overshoot the weight target. If your koji chamber is adequately sized to hold the rice spread out on cooling racks, it may just need supplemental heating. 200-500W of electric heat on a temperature controller should be sufficient.
Soaking and steaming - Use your normal setup for soaking and steaming, aiming to get the rice grains evenly hydrated and gelatinized through to the center. Soaking weight targets will depend on the rice type. Build in a rest period after draining to even out moisture. The post-steam (pre-cooling) weight target is 42-43%, however a wider range of 40-44% is acceptable. If the target is exceeded the rice can still be used, however the following drying phase becomes longer and more challenging
Note on Rice varieties - Koji Studies has relevant advice for home brewers on the topic of rice: “Even with the same water absorption rate, the quality of steamed rice varies depending on the quality of the rice and the degree of steaming. It is said that rice varieties grown for sake brewing more easily produce steamed rice with a hard outside and soft inside [gaiko nainan 外硬内軟] than general table rice, but that a similar quality steamed rice can be made with table rice depending on the raw material processing such as milling, water absorption, and steaming.” Section IV-1-(4): Steamed rice adjustment.
Non-sake rice varieties generally have a smaller and tighter grain structure with no shinpaku, so the corresponding steamed rice texture provides less opportunity to develop tsuki-haze koji. The advice above in Koji Studies gives a vote of confidence to aim for gaiko-nainan texture when using other rice types.
Hiki-komi - This is the drying step prior to inoculation, and potentially the most important part of the entire process. After steaming and cooling to 40℃ or below, the rice should be spread out on a drying rack and placed in a high temperature, low humidity space such as a warm oven (40℃, 50℃ max) or a well ventilated koji chamber set to 40℃. The goal here is to dry the rice over an hour or more, lowering the water content weight ratio to 28-31%. Expect another 1% drop from doing tane-kiri, so drying to 26-28% may be too low. When testing out a new rice type aim for 29-30% on the first attempt.
Note on Timing of koji growth to reach mori & naka stages
For the first 24hrs of germination & growth following inoculation:Each 1% decrease in the absorption rate at time of inoculation extends the growth time by 15 minutes. [Ref]
Each 1% decrease in the rice polishing rate (from 95% to 65%) extends the growth time by 12 minutes, or two hours per 10% decrease. [Ref]
Similar relationships exist with chamber temperature, tanekoji amount, and tanekoji variety.
Tane-kiri - Careful measurement of starter amounts and even distribution is important. For powdered starter the normal usage rate is typically ⅓ that of the granular type quoted in the article (the type commonly used by sake makers). Overall the starter amount ranges from 10%-50% of ‘normal’ rates, which for full strength powdered starter is 0.035g/kg up to 0.175g/kg of dry rice. The lower range is generally for making tsuki-haze koji, while the higher range will produce so-haze koji if the starter is fully viable. Playing with the starter amount is one of the key variables we can work with, since it will affect the final koji as well as the timing for the germination stage and beyond.
For powdered starter the quantity used is very small, so it’s best to extend the starter in rice or potato flour at a rate of 1:2 starter:flour or greater. The 1:2 ratio is convenient since it makes the strength on par with tane/granular koji. A 0.01g resolution scale is needed to measure amounts in the 0.10-0.50g range (dispersed strength) for a 1kg batch. For multiple tubs the starter amount can be multiplied for use across the full batch weight.
To distribute the starter make a custom shaker jar (篩 furui can). Use any small round container (metal works best to avoid static), and find a fine-spaced open weave fabric (silk, nylon filter cloth, stretch leggings) to cover the top and seal it with a rubber band. Shake the starter from several inches above the surface of the rice, and try shining a light from the side to view the spore cloud and gauge your technique. Shaking from too high will cause the spores float away rather than settle down in a normal drafty room.
As usual, break up rice clumps and turn over the rice a couple times during shaking to even out the spores. Expect to lose ~1% additional moisture from doing tane-kiri. Don’t worry about the rice temperature since it will rebound in the koji chamber.Momi-age / Germination - Pre-weigh each tub and mark the empty weight on it for easy reference. Sanitize the tubs and lids with a food safe sanitizer like StarSan or 70% alcohol. Place equal amounts of rice in each tub and record the weight(s). Loosely place lids or a plastic sheet over the tubs, allowing a small air gap. Set the chamber to 32℃*. Humidity is intended to be low, so don’t add any other water source to the chamber. All the moisture needed for germination is supplied by the rice and retained by the tubs.
* The original method calls for 35℃ ambient, however their temp chart shows a rice temp that is much lower. For smaller batches the higher chamber temp seems to speed up germination too much, and a lower starting temp may help induce tsuki-haze.Mori / Heaping - If there is any clumping break it up now, otherwise give the rice a quick stir to even out the moisture. If steps 2 & 3 have been done well, kiri-kaeshi is all but eliminated because the rice does not stick together much. Since we started out in tubs there’s no need to transfer between containers here. Record the koji weight.
Open the lid of the tub just slightly to allow for oxygen to reach the koji.At full scale, the tarai would remain covered with a plastic sheet at this time.
Naka-shigoto / Middle work - Surface growth starts to be visible and koji aromas increase. The amount of coverage may be 10-50% depending on how much starter was used, the moisture level, and the rice type. Mix lightly, breaking up any koji clumps so that the grains are loose again. Record the koji weight.
Shimai-shigoto / Finishing work - Mix again, making sure no grains are attached to each other. Once the koji reaches 40℃, or for simplicity at the time of shimai, the lid is removed and a clean dry cotton towel is placed over the tub. The towel can be folded to make 2-4 layers. This essential transition from covered to uncovered allows moisture to evaporate at a controlled rate while retaining some of the heat. Record the koji weight.
Monitoring and peak temperature - Only certain sake koji-kin varieties should reach 43°C or above, so 40-41°C is a reasonable peak target unless you are specifically aiming for a very high G/A ratio (gluco to alpha amylase). Once the peak temp is reached, the koji should be able to maintain close to that temperature until dekoji.
Because of the limited moisture content and low chamber humidity the koji has much less tendency to overshoot the peak temperature. If it does, you can easily correct by mixing and/or lowering the chamber by 1-2℃, but don’t remove the towel or the koji may dry out too quickly.
Note that with small batches the koji temperature will fluctuate so expect to see some variation here, especially after mixing. The goal is to maintain the koji within your target temperature range over this period to match the desired enzyme profile.De-koji - Decide ahead on a period of 10-18 hours after the target peak temperature is reached to time your dekoji. 10-12 hours is a reasonable starting point and should fall around 48-50 hours from inoculation. Longer times will develop more enzymes, which are generally used for shubo and hatsu-zoe koji additions. As home brewers we often use a single batch of koji for all additions, so we can try to match the koji strength with the desired brewing style and rice polish rate.
Remove the koji to a cool dry spot or the fridge, but don’t seal the tub right away to avoid condensation as the koji cools. Record the final weight. At de-koji the characteristic chestnut or mushroom (earthy) smell will be noticeable, both are acceptable results.
The dekoji weight can vary quite a bit when working with such small batches, so don’t worry too much about hitting a specific number. 12-17% is okay (even 10-20%). You can adjust downward further by drying/resting in the fridge. A higher dekoji-buai (weight ratio) will speed up koji dissolution in the moromi, and a lower ratio will dissolve slower.
Weights and temperatures by step
Targets for each step are marked in bold, so pay special attention to these before proceeding to the next step. Secondary conditions to be maintained in parallel during each step are also noted, these are less strict and can vary more.
Weight ratios and koji temperature targets are given at each step for consistency. Chamber temperatures are taken as constant throughout each step. The last column gives instructions on when to proceed to the next stage.
A spreadsheet version of the table above is also available.
* These narrow ranges are guidelines rather than critical control points
** Note that for less polished rice or higher starter rates, visible surface haze will often be higher.
Glossary
Weight ratio; moisture ratio - The current weight of the rice relative to the original dry weight. The difference is expressed as a percentage. This value is tracked throughout soaking, steaming, drying, and koji making.
Weight ratio (%) = [(Current weight of rice / Dry weight of rice) - 1] x 100%
Example: 1.3kg / 1kg - 1 = 0.3 x 100% = 30%.
Note that the weight ratio is not the same as true moisture content, since it does not take into account the starting moisture of the rice prior to soaking.
Results
Every rice type responds differently in koji growing. What the test group have found is that working with the same combination of rice and koji-kin can give predictable results by closely following the methods described here. Predictability gives the ability to adjust from batch to batch to further fine tune the results. As mentioned earlier, table rice varieties with a smaller grain size and with no shinpaku are much less likely to develop into tsuki-haze, but can absolutely become high quality so-haze koji with a desirable enzyme profile for sake.
Quantifying the quality of finished koji without enzyme testing or completing a whole brewing cycle can be a challenge, but one accessible measurement is to make a standardized amazake using the koji:
Koji test amazake: 1:2 koji to water ratio (eg. 50g koji, 100g water), hold at 55C for 3 hours, measure the Brix level with a refractometer.
Tips and Areas for further testing
Apply these suggestions to the steps above:
[Step 3] - Steamed rice drying - Homebrew sized koji chambers may not have the space or heating ability to dry the rice in a reasonable amount of time. If your chamber can’t reach the conditions for drying a larger heater could be temporarily added. Alternate options such as a food dehydrator or oven can work well too, so long as they have a low enough temperature setting to allow for slow and even drying.
[Step 4] - Tane-kiri - The choice of koji starter variety and the variability in distributing very small starter amounts can lead to changes in the growth rate and step timing, but this is easily manageable and becomes predictable over time.
[Step 5] - For momi-age you could try the original Ueda method of wrapping the rice in towels and plastic sheets, but at the smaller scale it shouldn’t make a large difference versus transferring straight into tubs.
[Steps 7-9] - Weight targets are not used here, but adjusting the chamber and koji tub humidity via ventilation and towels may be helpful to find a stable rate of moisture loss for your chamber setup. In general the humidity level needs to be higher than the original article calls for because of the higher surface area to volume of small containers.
You can make the effective humidity higher by leaving the lid partially covering the tub for longer into the growth phase. The goal is to have a slow and steady moisture loss in order to finish around 15% weight ratio at dekoji, but not much sooner since overly-dried koji stops producing enzymes.
Good luck making koji, and please share your experiences so that we can expand and improve on this guide.
Resources
Koji batch timing and weight log
A template for planning the timing of each step and keeping track of progressBoth of these documents may be copied for personal use after following the link.
References
Development of Koji Making Technique Using Water Absorption Rate of Steamed Rice as a Control Guide (Ueda-style Koji Making) and Its Applications. Dr. Morikuni Ueda, 2020, Journal of the Brewing Society of Japan, Volume 115 No. 2, p70-74.
agriknowledge.affrc.go.jp/RN/2010931636