5 costly drying process mistakes and how to avoid them

About the presenters

Dr. Susan Newman is director of Professional Services at METER Group. Her mastery of design thinking, engineering and moisture control has helped countless food and cannabis companies refine and optimize their drying processes.

Dr. Zachary Cartwright is lead food scientist at METER Group. He holds a PhD in food science from Washington State University and a bachelor’s degree in biochemistry from New Mexico State University.

Transcript, edited for clarity

Dr. Zachary Cartwright:

Hello everyone, my name's Zachary. I'm here today with Susan and we're here and excited to talk to you about five costly drying process mistakes and ways that you can avoid them. 

Whether you're baking or dehydrating, smoking or drying or curing – whatever your team calls it, or whatever that step is in your process – it comes down to one thing. You're removing water. A lot of food companies do this, and any food company that does will tell you that it's really hard to be precise and consistent. 

There are hundreds of different topics that we could talk about related to improving drying, but today, we'll focus on the top five mistakes that we see. 

The mistakes we’ll talk about are:

1. Misunderstanding the correct measurements that you should use, 
2. Using the wrong method for that measurement, 
3. Letting operators work on intuition, and maybe even using their hands to guide their decisions, 
4. Sampling in the wrong places, and 
5. Leaving your control loop open. 

What’s at stake?

Before we look at the five mistakes, I want to talk about how much is at stake, the cost of inaction, and what it costs companies when they don’t focus on reducing their moisture content variability. 

We have a figure here that shows the cost of inaction. On the X axis, you have time in weeks, and on Y axis, you have cost. 

This example is for different pet food products. You'll notice that every week that goes by costs tens of thousands of dollars. We see this not just in pet food, but in lots of other products as well. 

A lot of companies aren't even aware of what it's costing them to have a high degree of variability in their moisture. 

What have you seen, Susan, in your experience when you go and talk to clients? Are they aware of these figures and what's it costing them?

Dr. Susan Newman:

Great question. They are aware, but they are under-aware. They underestimate the cost. 

This chart comes from real client data – before we worked with them, they were underestimating the cost by about 80%. It was huge. 

They knew there was an opportunity for increasing yield, but they thought it was going to be pretty small, because data can be challenging to find and interpret in the food industry. A lot of times it lives in a paper record. So really digging in and putting in the energy helped them see that they were losing almost eight times more than they thought.

People often underestimate where they can be by controlling the process. This is an example from pet food, where the cost of ingredients is quite low, but we see this same situation with customers that are using whole muscle meats, for example, or cannabis – where the cost of ingredients or product is much higher. 

the cost for ingredients in pet food could be cents. But compare that to cannabis, where you're looking at, depending on the market, $2,000 per pound. Or meat – $8 per pound is pretty average for a price there. So this is exponential for those markets.

Mistake #1: Misunderstanding the correct measurements to take 

Dr. Zachary Cartwright:

That brings us to our first drying mistake. What I notice among a lot of food companies is that they misunderstand the correct measurement they should use. 

Basically, water measurements can be summed up into two types: moisture content or water activity. I want to take some time to note the differences between these two measurements. 

Before I compare them, what do you see, Susan, when you go talk to a client? Are they using one of these or both of these measurements? How are they being implemented?

Dr. Susan Newman:

I see everything. You can't even imagine what I've seen. From NIR to moisture balances to Karl Fischer. There’s a lot of water activity as well, but different industries have different required CCPs. Pet food, for example, is heavily into moisture content as their CCP. 

A lot of companies we work with understand the importance of water activity for safety parameters, so they're using that a little bit more. I love seeing that. We have a long way to go, though – I see a lot of people using moisture balances out on their factory floors. I see a lot of water activity typically in the lab, not so much on the production floor.

I had one client that – I love telling this story because it's so fun – I met with a CFO, and he was super excited about moisture content. So in their lab you would see 10 moisture content devices and one water activity device. 

When I met the QA team leader – her name is Theresa, I love her – she just loved this one water activity meter, because she understood where pathogen growth occurs. She understood that bacteria won’t grow under 0.8, no molds or yeast grow under 0.6, and nothing grows under 0.6. 

She also understood that water activity is the driving force for crispiness. If you're making a crunchy product, you need water activity in your life, and she understood that. 

The CFO was really leaning into moisture content because that is the yield parameter – and he's right. If you're running a business, you need to understand moisture, but you also have to make a safe product. understanding both of these two measurements and using them appropriately is somewhere where we really need to get in food.

Dr. Zachary Cartwright:

We have a table here that  really helps to summarize the difference between these two measurements.

Water activity is a measure of energy. It's a thermodynamic principle. That energy is important because it helps us know if certain chemical reactions can occur or if microorganisms can grow or if texture can change. Moisture content is simply an amount. So if you were able to remove every molecule of water from a food sample or from a product, that would be the moisture content. That is extremely difficult to do, to remove all of the water and get an accurate reading. 

Water activity is qualitative. What I mean by that is, even though we get a quantitative number, it's qualitative because we can relate that right back to safety and quality of that product. I see a lot of companies try to take a moisture content measurement and relate it back. But the inherent variability in moisture content makes it extremely difficult to do that. Moisture content is more quantitative, just like you said. A CFO or a decision maker thinking of how much this is going to increase their yield or revenue is going to be interested in moisture content. 

Water activity is a driving force for chemical reactions. Moisture content, even though it has a relationship to chemical reactions, it's really hard to understand the relationship. Water activity is much higher precision and accuracy in that regard, so it allows us to make the determination and relate it back to the safety and quality. With water activity, we have known standards. These are different salt solutions that have the same water activity every time, so it makes it really easy to verify the calibration of a water activity instrument. Moisture content doesn't have these types of standards. There's nothing that has an inherent moisture content that we can compare to, so it's really hard to know if you're getting an accurate number or not.

Finally, water activity is unitless. It ranges from zero – something that has no energy – all the way up to one, something that has the same energy as pure water. When we talk about moisture content, this is generally expressed as a percent, either on a wet basis or a dry basis. 

Hopefully this table helps you to understand some of the differences between these two water measurements.

Mistake #2: Using the wrong measurement methodology

Dr. Zachary Cartwright:

Let's move into the second mistake: using the wrong measurement methodology. 

Whether you're using water activity or moisture content, there are different methodologies for each of these, and it's really hard to control a process if there's a lot of standard deviation and a lot of variability inherent in the method that you're using.

We have a graphic here that shows results for one product, tested with different methodologies, and the results from each of those methodologies. On the X axis, you have different methodologies. On the X axis, you have moisture content and water activity. 

The first thing I want you to note is that even though this is the same exact product, that moisture content is all over the board. But if you look at the water activity, it is very consistent – almost no change in water activity, even though the moisture content measurement is giving us a high degree of variability. 

Why is this important, Susan?

Dr. Susan Newman:

When we did this study, I was so excited. We took a product, brought it all to a point where it was acclimated and perfectly consistent, ready to go – so you'll see that water activity is a very consistent 0.42. 

We sent it off, then looked at different moisture content readings. I was so shocked to get these results. The lowest we got was 3.5, and the highest is almost 10 – from the same product. 

I tried to put myself into a QA team's shoes. How would they feel if they got these results? What actions could they take from these? When I see results that are so varied, I feel defeated for them. 

Say you only send this out to one lab, so you only get the 3.5, what do you do? You think that you are over-drying your product, so you add more moisture, reduce temperatures, maybe decrease cook times, which can mess with your CCP kill steps if you’re not careful.

Then on the other end of that spectrum, if you get a reading back of 10 and your product can be up to 12% – and this product was – you get that and you think, "Wow, I'm doing a really good job." 

But looking at all of this data, I was just shocked at the variance that we saw in this and how hard it is to really do a good job every day to make your product consistent.

Dr. Zachary Cartwright:

We recently had a webinar on moisture measurements, and something that stood out in that webinar once we had collected some data was just how much variability there is, especially for moisture balances.

This goes back to what you said earlier. A lot of facilities that we serve might have 10 moisture balances and one water activity meter, but that moisture balance can have a variability of 4 to 7%. That's really common, and can be because of a different person using it or a different methodology or so on, and this graphic does a good job of showing that variability. 

What would you do if you were that QA person, Susan? What would you do after seeing this data? How would this inform you to make a decision about the method that you should be using?

Dr. Susan Newman:

Seeing these individual data points would be huge for me. I've always known that water activity is critical for us to understand due to its accuracy. I do see moisture content varying with my clients, plus or minus 2% is across the board, but even up to four to seven like you said. 

If I saw this data, I would be shocked and really start questioning my methods. I’d think, “How can I do this in a better way? What is the truth? How can I lean into more accurate measures such as water activity, but still apply them to moisture content, which is a CCP, that’s critical to my yields and profits?”

I would really want to lean into this water activity, but understand the difference and the relationship between the two and there is a relationship. That's where I would spend my time. But man, if I just got one of these data points, I would think I was doing awesome, or I would think I needed to be fired.

Dr. Zachary Cartwright:

It's important to note that there is a relationship between water activity and moisture content, and this is something that we're going to touch on in the next mistake. 

Mistake #3: Letting operators work on intuition 

Dr. Zachary Cartwright:

Let's talk about the third mistake: Letting operators work on their intuition instead of giving them real-time, accurate data to help them make informed decisions. 

You have some good stories about this, Susan. I'll hand this over to you.

Dr. Susan Newman:

I spend most of my time on mistake number three. There are a lot of operators, especially ones that have been working for a long time, who keep tribal knowledge. 

One of my favorite customers – I hope he's watching today. Hi, Harley! – was working on a line where they were making a dog biscuit. They made a dough to start, and he was convinced that he could reach into the dough, feel it, and tell if it was going to run well or not. I challenged him on that, because really what he was feeling was temperature. As you die roll, colder product comes through a little bit nicer, but hands are not calibrated to do that kind of work. If you're relying on that, call me. We don't want to do that, but that’s really common.

Even QA was on board with this because that wasn't a point in the process where they were measuring anything. It didn't really matter as a CCP. So for him, intuition was it. But how is he going to train people? How do you train people to have this sense of feeling? Working on intuition: not always the best idea. That's my first example. 

Another example that I see oftentimes is in cannabis. Historically, there wasn't a lot of science in cannabis. There was a lot of snap testing. If a customer's doing a snap test, about 0.4 water activity is where they're going to get that feeling. I've done a lot of testing on this –  that feeling of a snap on the branch happens at 0.4, and that's a problem because at that point, you've already degraded terpenes in your product.

We know that at 0.5 water activity, terpenes start to degrade in cannabis. If your average is coming out the door at a 0.4 water activity, you've already degraded terpenes. You'll see this same thing in chips and other things. If you over-dry, you start losing quality.

In this graph, the red zone is the ideal. This graph specifically is for cannabis, this is a cannabis isotherm showing the relationship between moisture and water activity. I want all of our clients to hit that beautiful sweet spot. That's the point where we're safe from pathogen growth. We don't have much growing on cannabis under 0.625. The one we have to look out for is aspergillus enchiladas, that can grow down to 0.625, but I still want them to have this beautiful yield.

We want to keep them between 0.55, for terpenes that are nice and beautiful, and then an upper limit of 0.625. By hitting that, they're going to have a beautiful product, but also the most profitable product to sell as well. 

Based on this relationship between moisture and water activity, a snap test typically comes in at about 8% moisture. If you're relying on a more accurate method, such as water activity, you're going to be able to get this precision that you need to be able to dial it in here and up to 12% moisture. For cannabis, a 4% increase in moisture, even if you're doing 50,000 pounds a year at a sale price of 2,000 a pound, you're looking at this being a $4 million problem by optimizing that 4%.

That's an extreme case because of the value of the product, but we see this all the way down to even our low value products where this is happening. Another really cool thing about understanding the isotherm – the relationship between moisture and water activity – is you can see the reason Harley wasn't very effective. You can see that this large change in water activity is only a small change in moisture, so moisture doesn't really change that much, but water activity changes a large amount. That's why hands don't work. That's why relying on water activity as your measurement of choice here is the best.

Dr. Zachary Cartwright:

It's important to understand that this moisture sorption isotherm we're looking at, this is  something that METER Group specializes in. We have a patented technology that allows us to either make a desorption curve and understand how water is removed from a product, or an adsorption curve to understand how water binds to that product. Then using that curve and setting a target for your water activity, we can find a sweet spot and please everyone. We can make the QA person that understands the science happy, so they know their target and also understand what that moisture content target should be.

This graphic does a good job of summarizing everything that we need to know about the water in that product, making the scientists happy, as well as the decision maker, and understanding what that sweet spot should be. This is something that we can make for every different type of product. Even every formulation may have a different curve that we need to consider.

Mistake #4: Sampling in the wrong places

Dr. Zachary Cartwright:

Let's move into mistake number four: Sampling at the wrong spot. 

Susan, when you go into a food facility, where do you see people sampling and where should they be sampling?

Dr. Susan Newman:

Awesome question. I typically see people sampling post-cooler. So in this example, before it goes into a package, that's typically where CCPs are, so we'll see a lot of heavy sampling in that area, and then some sporadic within the process.

I've got some clients that are starting to understand the variance in ingredients and how that affects the dough. In this cookie example, we've got a dough area, we're doing the mixing of those raw ingredients, we've got an oven, and then we've got a cooler. In the mixer, the dough is typically 0.99 water activity.

Then at the cooler, for a cookie to get that nice crunch that you want, that's about 0.3 water activity. By sampling those two areas, that gives you one view. But the part that's missing is the most important part, the part that takes the most energy to run: the oven.

What I love to see is clients that are branching out and really understanding a target for each process. Dough is one process, the oven is one process, and then cooling. We could even expand this also and look at how you're storing after you cool it but before it goes into the package. So depending on where you're at, at 0.3 water activity, if you've got a really dry environment, you could lose moisture. In a humid environment, you might gain moisture and push levels to where you've got gross cookies that aren't crunchy. Understanding each of these processes is really important.

Dr. Zachary Cartwright:

Something that's missing from this graphic is that a lot more companies are measuring incoming ingredients. A lot of companies now will set a spec, maybe plus or minus 10% water activity of incoming ingredients. The reason for that is that any variation in your incoming ingredient can be carried through to that final product. 

That's becoming more and more common. Have you seen that in your experience with some companies?

Dr. Susan Newman:

Yeah, that's why I'm over here smiling. One of my favorite clients is Sunrise Fresh. They have a family farm. It's awesome, so they're growing cherries, walnuts, and things like that. 

One of their products that is most troublesome is the cherry. From year to year, depending on the amount of water and sunshine the cherries’ brix levels can change. So now they look at the cherry isotherm to understand how that change in brix affects the cherry. If you're drying it and putting it in a bag to munch on, that's one thing. But if you're using it as an ingredient in another product, like an energy bar for example, then it has a bigger impact, because sugar is such a strong humectant. 

That's a fun way to look at ingredients. A small change from year to year variation in a cherry can be a big change in a granola bar.

Dr. Zachary Cartwright:

It is important to note that we can make an isotherm, whether it's every season or every formulation, and that isotherm really helps us to pinpoint the sweet spot. We can make that isotherm for every product or formulation and then use that information to set the target correctly. 

Something else that we could note about this figure is that I notice some companies using benchtop water activity meters to try to get an in-process measurement. That can be somewhat effective, but if you are pulling a sample, then having to wait for the temperature to come to the right temperature for a reading, or if you have to do a lot of sample preparation, it might take you 20+ minutes to get a reading. By that point, it's already too late. The changes have already happened, and you can't correct effectively. 

Why is it so important to be able to take an in-process measurement, Susan?

Dr. Susan Newman:

I know exactly what you're talking about. Some people think that if we don't know, ignorance is bliss, we can keep running. But they’re running product that is out of spec. Maybe it hits spec for CCP, but it doesn't have the quality that you want. 

By understanding exactly where you're at and quickly getting that feedback and making these micro adjustments you can make it perfect, and that's where you want to be. 20 minutes is too long to wait. We're running belt speed as high as we can go because we want that throughput, so 20 minutes can be a huge amount of diversion and rework. I like to see this instant-by-instant. And if you can close the loop, even better. That's something we're going to talk about next too.

Mistake #5: Leaving your control loop open

Dr. Zachary Cartwright:

All right, so our fifth and final mistake is leaving your control loop open. 

What do you mean by this, Susan? You touched on it earlier in mistake #4, but what do you mean to have an open loop versus a closed loop?

Dr. Susan Newman:

Pandemonium is what I mean. By leaving your control loop open, that means that you're just running product. You probably know the time and temperature recipe for a cookie or whatever your product is. Leaving the loop open means that you're setting those parameters and you're letting it run until there's a problem. You're not really optimizing. 

Where I really want to see clients thriving is at that point of variation reduction. I want them generating a product that is really tight to their spec.

You'll see in the graphic here, this is an example of our SKALA Solo system running, and in this we're able to close the loop on a process instead of letting it run. 

Variability is the enemy. If we have a lot of variability, we've got a lower yield, we've got a lower throughput. It's costing more to run those ovens as well. Plus a lot of rework, expensive people, and then where are you going to hold it? You'll have to hold this product, and then when you hold it and it's out of spec, it can lead to mold issues. Then it's something that you have to throw away, and your customers aren't going to like it either. They don't want a moldy cookie. 

By closing the loop and understanding this data, we've got a historical record. A product like SKALA (or others on the market) can give you that historical data, but understanding what it should look like and making changes quickly, you can't do that if you're a human. You have to do it by looking at data on a second-by-second basis, and then knowing precisely when to make an adjustment, so you're not throwing your system out of balance.

Dr. Zachary Cartwright:

To clarify, that open loop system is still dependent on a human – of course there's going to be error associated with that. The closed loop system is still being monitored or watched by an operator, but quick adjustments are being made by the instrument itself or by the machine. 

Maybe you can talk about that a little more.

Dr. Susan Newman:

Certainly. SKALA Solo looks at temperatures within the system – so inlet, outlet, and product temperature – and then we've got a patented algorithm. We understand in-process moisture and water activity, and we collect both, but I like to train around water activity. 

You'll see on this PLC example, there's a spot here that says the water activity target. And what we're looking at here is, I want this to run on that target, and then operators are going to take a sample to verify that the algorithm is running efficiently. It's making adjustments all the time, micro-adjustments that don't have a huge impact on the overall process that you validated with your R&D department.

We want to look at second-by-second data, and every minute making the decision to make a micro adjustment. And not just that, you'll also see here that there's multiple zones. You’ll be able to see which zone has the biggest impact on the outcome that you're trying to achieve and adjust in that zone.

We've talked about water activity and moisture content, but now we're bringing this all together, looking at historical records and how we can really bring industry 4.0 data to make real-time, fast, actionable outcomes for operators.

Dr. Zachary Cartwright:

One way to understand this is, if we allow this PLC animation to play, you'll notice that there's a certain target and it's lower than optimal and has a lot of variation. But just 30 minutes into implementing SKALA Solo, you'll see that variation reduced, and then you can  increase the target. 

This goes back to our very first graphic of what's at risk here, or how much does this cost companies that are unable to make the adjustment. This graphic does a really good job of showing how powerful SKALA Solo is to very quickly reduce that variation.

What product is this for, Susan?

Dr. Susan Newman:

This is an extruded pet kibble, dog food.

Dr. Zachary Cartwright:

From here, I want to move on and ask about operators. How can you have a brand new hire be just as good as that person who's been in the industry for 30 years? Does this tool allow that to happen?

Dr. Susan Newman:

Yeah, nothing is magic, but this tool definitely takes decision making out of a human's hands and does it for them. It's going to make these adjustments in a closed loop system, connected straight to a PLC. 

So it gets PLC data, which could even be at the micro level, that PLC data comes in, it reads that, understands those inlet, outlet, product temperatures, and then automatically adjusts. 

Now, from device to device, startup and shutdown is always going to be difficult. We help navigate that as well by helping operators to really focus on how they do a startup and a shutdown, and then we can take it from there, so essentially putting the process into cruise control.

Review and wrap-up

Dr. Zachary Cartwright:

All right, to wrap things up, I want to move through the five costly mistakes one last time. 

The first one was misunderstanding the correct measurement that you should be using. The second one was using the wrong methodology for the measurement. The third one was letting operators work on intuition – and we went through a couple stories that Susan had. The fourth one is sampling in the wrong place – how we need to not just sampling the final product, but sample throughout the process. Then the final one was leaving your control loop open. 

If you have any questions about any of these topics, or questions about water activity, moisture content, isotherms, or SKALA Solo, please reach out to us, visit our website, or reach out to Susan or myself directly. We'd be happy to take your questions. 

We do have a couple questions here now, and we'll go ahead and open the floor for those questions and answer as many of them as we can.

Q&A #1: Can you speak a little bit about the differences and commonalities you’ve seen in the way different products dry? Is a product-specific drying approach always necessary?

Dr. Susan Newman:

Pathogens are pathogens, and where they grow is really specific to that pathogen, not the product. We know bacteria are not growing below 0.8, yeasts and molds below 0.6, so that's the commonality. 

The difference is really understanding the product, and we talked about looking at the isotherm, that difference between moisture and water activity is different for every product. The way that you get to a safe point is going to be different for all those products and specific to that product. 

The difference really is understanding where to go fast and where to slow down. For a cookie, it's a quick process. You don't have to worry so much about microbial growth. When you're working with a product like a salami, that might be a 30- or 90- day process. You want some bacteria to grow in there, but have to do that in a safe way as well. 

So across the board, these differences exist. In cannabis, if we go too fast, we kill the terpenes on the outside of the product and leave the inside moist. So in terms of speed, that's really where the differences lie.

Dr. Zachary Cartwright:

It's a good point that even though the targets may be the same for some of these different products, the way that these products lose water or even gain water is really product specific or even formulation specific. But using an isotherm, this is something that we can understand. This can help you set your targets. It can also depend on the methodology you're using, whether it's a spray dry method or batch oven or something else, we are able to look at each of those methods and help you to reduce variation and hit your target.

Q&A #2: What advice do you have for someone trying to dry high moisture products (60% MC or greater) as quickly as possible?

Dr. Zachary Cartwright:

Well, generally products with that high of a moisture content are things like dressings or maybe jams or something that are much higher in moisture. If you're working with dough or something like pet food, you need to be careful of drying too quickly, causing encasement. This is where the inside of the product has a higher moisture content, higher water activity than the outside, and once the product reaches equilibrium, you start to have problems with microbial issues. 

I'm not exactly sure what the product of interest is here, but I will say that for certain products, you do need to be careful not to heat it too quickly. 

What's your experience, Susan? Is there something I'm missing here?

Dr. Susan Newman:

We need to look at hurdle steps here. I'm interested in what humectants that you might be using to help reduce that water activity for more of a stable, safe product, so you don't have bacterial growth. Maybe pH, so you might look at pH as a part of your hurdle technology as well, to make that a little bit safer from pathogen growth.

Q&A #3: I work with snack foods. In terms of SPC, what attribute should I use to measure and control to reduce process variation?

Dr. Susan Newman:

We've talked about this quite a bit. Looking at process variation, you have to have a really good way to measure, a great yard stick, and you do that through accuracy. You have to look at water activity, where you can get accuracy of 0.003, 0.005, which is in our TEs and our AQUALAB 3. Moisture balances, we showed today, saw readings all over the place from 0.35 all the way up to almost 10. 

If you're looking at statistical process control in your process, you have to look at a really great way to measure, and that's going to be water activity. That's going to be the attribute that you want to establish for each step of that process, from mixing to cooking and cooling, and before packaging.

Dr. Zachary Cartwright:

This goes back to isotherms again, a lot of snack food products are in that sweet spot we've been talking about. A 1% change in moisture content can be a 30 or 40% difference in water activity. It makes sense that you want to use the measurement with the highest degree of precision and accuracy, and that's going to be what activity every time.

What parts of the drying process are most likely to promote microbial growth? What can be done about it?

Dr. Susan Newman:

We looked at the cookie example, and I've seen issues where people using a batch process, mix all their ingredients, add their water, then it gets pressed or formed, goes through the oven, then they put in more ingredients, but they haven't done a thorough wash-down. 

The first part of this is really making sure that you have a great sanitation program, making sure that you're washing down as frequently as you need to, swabbing to see if you've got growth, and logging where those high growth locations are located. You'll see that typically in places that are hard to get to. The problem is, they're hard to get to – so people don't clean them.

I do see that quite a bit. What can be done about it? Looking at your programs, looking at your shift change over, that's really where I'm leaning to, as well as getting your product below spec for microbial growth. Unless you're using a kill step in there, you're not killing anything, but you want to get it down to a point where they’re not growing and multiplying. 

So again, bacteria, 0.8 water activity, yeast and mold – we're going to pound it into your head – 0.6 for those. Getting to those limits as quickly as you can and maintaining them there.

Dr. Zachary Cartwright:

The only thing that I'll add to that is even if you get below those limits, keep in mind that the conditions of your facility can promote microbial growth if humidity or temperatures are high. Even if you are below the limits that Susan mentioned, you can be in an environment where the microorganisms are already there, and then once the environment is allowing for growth, you're going to see a bloom, the microorganisms will start to grow. 

So keep that in mind, even if you control throughout the drying process, that product, if it's close to a limit, still needs to be monitored very closely.