A cocoa farm generates roughly one tonne of pulp for every tonne of dried beans it produces*, and almost all of it is discarded before the beans ever reach a chocolate factory. A recent Indonesian study asked what happens if you remove 40 percent of that pulp deliberately, add a defined starter culture, and turn the removed material into a separate product. The answer changes how we think about fermentation as a controllable process rather than a fixed ritual.
Cocoa fermentation looks deceptively simple from the outside. Fresh beans go into a box, the mass gets turned periodically, and several days later the beans come out ready for drying. Inside that box, a complex microbial succession is doing the real work. Yeasts, lactic acid bacteria, acetic acid bacteria and other organisms interact with the pulp surrounding the beans, consuming substrates and generating metabolites, heat and acids that ultimately build the flavour precursors inside the bean, focus on the word, precursors.
That raises a question worth asking directly. How much of that pulp do you actually need for a successful fermentation?
The Pulp Is Not Just Packaging Around the Bean
Cocoa pulp provides both a physical environment and a nutrient supply for microbial growth. Yeasts ferment the carbohydrates in the pulp, producing ethanol, carbon dioxide and heat. Lactic and acetic acid bacteria then work through sugars, ethanol and citric acid, producing the metabolites that carry the fermentation forward and shape eventual flavour.
But more pulp is not automatically better. Excessive pulp can increase acid production and push fermentation toward sourish beans, and earlier research had already shown that fermentation could still proceed with a reduced quantity of pulp. That leaves an interesting opening. If some of the pulp can be removed before fermentation without harming bean quality, the removed material stops being a disposal problem and becomes a second raw-material stream.
Related reading: The Importance of Viscosity in Chocolates Part 2: Why “Correct Viscosity” Still Fails in Industrial Production
The Experiment
Researchers worked with healthy bulk cocoa beans of the Lindak variety from smallholder plantations in Gunungkidul Regency, Yogyakarta, Indonesia, and ran three fermentation treatments side by side.
- Treatment A, Depulping: beans partially depulped, no starter culture added.
- Treatment B, Depulping plus starter culture: partially depulped beans received the Lactiplantibacillus plantarum HL-15 starter culture.
- Treatment C, Whole beans: the conventional comparison, whole beans with no starter culture.
Pulp was reduced by 40 percent using a depulping machine. Fermentation ran in wooden boxes with roughly 40 kg of beans per batch, covered with banana leaves, for five days, with turning every 24 hours. Temperature and pH were tracked at multiple points inside the fermentation mass, and microbial populations were analysed throughout the process rather than just at the end.
What Happened to the Beans
The reduced-pulp treatments did not prevent good-quality beans from developing. The combination of depulping and the L. plantarum HL-15 starter culture produced high-quality cocoa beans with a strong sensory profile, and the starter culture also helped regulate the microbial consortium and suppress fungal growth.
There is an important nuance sitting inside that result, though. The fermentation index came out at 1.6 for the conventional whole-bean treatment, 1.3 for depulping alone, and 1.5 for depulping combined with the starter culture. So the finding is not that less pulp produced a better fermentation. The finding is narrower and more useful: a substantial pulp reduction did not prevent high-quality fermentation under the conditions tested. That distinction matters if you’re trying to use this research to plan an actual process, rather than just quote a headline number.
Why Add a Starter Culture at All
Traditional cocoa fermentation is largely spontaneous. The microorganisms already present in the cocoa mass develop and succeed one another without deliberate control, and that spontaneity introduces variability from batch to batch, farm to farm, season to season. Controlling the microbial community with a starter culture can offer faster fermentation, more controllable conditions and more consistent bean quality compared with leaving the process to chance.
The researchers chose Lactiplantibacillus plantarum HL-15 because it had already shown good cocoa bean quality and additional resistance to mould growth during earlier work. It was added to the partially depulped beans before fermentation at a concentration of 10 to the power 10 CFU per mL, at 0.1 percent by weight. In this experiment, it helped regulate the microbial consortium and suppress fungal growth, which matters because fermentation control is ultimately biological control, not just a temperature and timing exercise.
Related reading: We Don’t Really Manufacture Chocolate. We Manufacture Its Structure.
The Process Hiding Inside the Experiment
This is the part that interests us most from a manufacturing perspective. Removing part of the pulp creates a second raw-material stream that did not exist as a usable product before, rather than simply eliminating material from the process.
On most farms, cocoa pulp is treated as a residue, and its disposal can contribute to odour problems and soil pollution nearby. The researchers investigated turning the extracted pulp into a probiotic drink instead. From 40 kg of wet cocoa beans, the process generated approximately 4 kg of pulp, enough to produce around 20 litres of probiotic drink. The study’s value-added analysis estimated that those 4 kg of pulp could generate IDR 262,500 in added value.
Partial depulping stops being just a fermentation intervention at that point. It becomes a question of process integration, where the pulp removed from one process becomes the input to another.
Wondering whether a by-product on your own line could be worth more than the cost of discarding it?
An Industrial Audit reviews your full process, inputs, outputs and everything currently treated as waste, and shows you where the real value is sitting.
The Economics Are Not as Simple as Waste Becoming Profit
This is where the research deserves careful reading rather than an enthusiastic headline. All three fermentation methods produced similar output values in the economic analysis, but the depulping-plus-starter-culture method carried higher costs because of the starter culture itself. The study reports a cost of IDR 2,550 per kg for method B, against IDR 50 per kg for method A and no additional cost for method C.
At the same time, the whole-bean method had the lowest yield, at 35 percent, against 37.5 percent for both depulped treatments. There is no simple story here where removing pulp automatically converts into extra margin. The economics depend on what happens to the removed pulp afterward, the added processing cost of any starter culture, and the market value of whatever product the pulp eventually becomes.
Related reading: The Yield Loss Tax in Chocolate Factories
What This Means for Cocoa Processing
There are three separate ideas worth pulling apart here, because they get bundled together too easily.
- Fermentation can be engineered rather than left to chance- A defined starter culture introduces a level of control that spontaneous fermentation cannot offer, shifting the objective from letting fermentation happen to actively influencing which organisms dominate and what metabolites they produce.
- More substrate is not automatically better – Pulp is essential to microbial activity, but the research shows that excess pulp can push fermentation toward increased acidity. That raises a genuine process question: could the pulp level itself become a controllable fermentation parameter, rather than whatever quantity nature happens to leave on the bean?
- A by-product can become part of the process design – Once part of the pulp is deliberately extracted, the relevant question becomes what happens to it next. The study demonstrates one route, a probiotic beverage, but the underlying principle reaches further than that single product.
Framed that way, the process looks less like cocoa pod turning into beans plus waste, and more like cocoa pod turning into beans plus a second product. That is a fundamentally different way to look at the same raw material.
What the Study Does Not Prove
This matters if anyone is tempted to treat this as a ready industrial blueprint. The experiment used one cocoa variety, one starter culture, one set of fermentation conditions, and a relatively small batch size of around 40 kg in wooden boxes over five days.
The paper does not establish that every cocoa variety tolerates 40 percent pulp reduction, that 40 percent is the optimum depulping level, that starter cultures always outperform spontaneous fermentation, that the method transfers cleanly to industrial-scale fermentation, or that probiotic beverages made from cocoa pulp will be commercially viable everywhere. Those remain separate engineering and commercial questions, and each one needs its own testing before capital gets committed to it.
From Fermentation Box to Process System
For anyone looking at cocoa processing from an engineering perspective, this is the useful takeaway. A fermentation box holds more than beans sitting for five days. Underneath that stillness runs a biological process moving through a chain: substrate, microorganisms, temperature, pH, metabolites, flavour precursors, bean quality. Change one link in that chain and the others respond, whether or not anyone intended them to.
This study changed two links in that chain at once, reducing the pulp available to the system and introducing a defined microbial starter culture, while also asking what could be done with the material removed from the system. That combination, treating fermentation as a system with measurable, controllable and potentially monetisable parts rather than a fixed traditional procedure, is where the next generation of cocoa post-harvest processing is likely to come from.
Related reading: Batch vs Continuous Tempering: The Difference Nobody’s Actually Watching For
The Key Takeaway
Whether cocoa can ferment well with less pulp is not really the interesting question anymore. This study suggests it can, at least under the conditions tested. The more useful question for anyone running or planning a cocoa operation is broader: can fermentation be designed as a controlled process where pulp level, microbial population and the utilisation of whatever gets extracted are all considered together, rather than handled as three unrelated decisions?
That is a bigger question than most fermentation research sets out to answer. This study is one experimental data point toward it, not a finished blueprint. But it is a useful reminder that the parts of a process labelled as waste are usually just the parts nobody has measured properly yet.
Work with Rudvik Engineers
Curious what an audit would actually find on your own line?
An Industrial Audit maps your process end to end, fermentation through packaging, and flags exactly where value is currently being discarded.
Planning a cocoa or chocolate processing line and want the full picture before you commit capital?
An Industrial Technical Discovery Session walks through your formulation, process flow and by-product streams in one focused session.
*Note:- While fresh pod wet mass creates roughly an equivalent ratio of pulp to dry bean output, precise extractable yields range between 500–600 kg (approx. 150–200 liters of juice) per tonne of dried beans. Additionally, only a portion of this pulp can be harvested prior to fermentation without compromising bean quality.