Skip to content
Feb 24 / scott.campbell

Water Activity and Candy Texture – Hard Hearts

Conversation Heart in AqualabI have always been partial to Sweethearts: chaulky, heart-shaped chunks of candy sold in a box that could alternatively be used for a deck of cards. Perplexed by the needy language that adorns the hearts, I’ve always been content to just munch away on handfuls of them and get my the pastel-colored sugar high.

Recently, though, my work colleagues and I have looked at how texture of products like this is related to water activity. For example, the Sweethearts, as manufactured, start out as a dough that’s shaped into hearts and then baked/dried for 30 minutes. Having never visited the factory, I’m not sure what the water activity is when they’re done. But I’ve had both crunchy (think dry Lucky Charms Marshmallows) and soft. We wanted to know what the dividing line between the two textures was in terms of water activity, and which texture we liked the best.

When we bought some Sweethearts at the store, they were found to be 0.330 aw, and crunchy. After sitting in our bone-dry food lab for a week, they had declined to 0.18 water activity units, and had the mouth feel of ice shards. We then equilibrated the candy chunks over different salt solutions. This brings them to a known water activity, after which we ate some more of them. The verdict: all of us like the softer candies that had a water activity of 0.65 – much higher than the water activity of the product we bought off the shelf. Note that if you live in a rather humid environment (above 60%) you’ll probably find your Sweethearts to be soft, while in the arid farmland of eastern Washington state, ours are usually crunchy.

Where does this leave Sweethearts? With not enough moisture. I recognize that, as a novelty item, Sweethearts may fall into the same category as baseball card gum – the fact that they’re edible is more important than what they actually taste like. Still, the moisture transfer issue is something that Sweethearts scientists must be aware of, since I’ve seen packages covered in plastic wrap before. Our conclusion is this: water activity is the driving force for moisture migration, and moisture migration determines whether we get a nice texture or an unpleasant one. Sounds like a lesson for customers with hard hearts, soft taffy, and everything in between.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Dec 23 / scott.campbell

Bean There, Done That

There is no substitute for experiencing a product first-hand.  In our case, that meant mixing up 5 batches of ice cream to test whether I’d been ripped off by some vanilla beans I bought on Amazon.  Per my previous post, when I ordered a bag of 20 vanilla beans from Amazon for $25, it seemed too good to be true.  The first sign of trouble with these beans was their water activity: 0.98 upon arrival.  Although I dried them out (down to 0.88 water activity units), which prevented mold, but then the beans didn’t seem to add any flavor to the products where they were supposed to be used.

For my next step, I ordered some “real” vanilla beans.  I’ll call them “real” because the ones I got were similar in appearance and texture to the vanilla beans I got from Costco, and have loved in the past.  They also cost from $3-4 per bean.  I cut up one to test its water activity when the beans arrived.

photo (13)

No contest: the real beans had a far lower water activity value of 0.86.  This indicates that these beans are a totally different product than the Amazon beans, which were soaked in something strange before then came to us.

I decided to make 5 batches of ice cream, all with the same recipe.  The only difference would be the vanilla, and I would either use 2 tsp of vanilla extract, or 2 vanilla beans, when called for in the recipe.  The recipe is here – an Alton Brown classic.   The five batches would be:

  1. No vanilla at all
  2. Vanilla extract
  3. “Real” vanilla beans
  4. Amazon vanilla beans, as packaged (wet)
  5. Amazon vanilla beans, dried to a safe water activity

As I was making the ice cream mix, I immediately noticed two things about the vanilla beans.  First, the real vanilla beans smelled totally different than the Amazon ones.  Second, the dry vanilla beans didn’t really resemble the real vanilla beans at all.  They had about the same pliability, but the inside of the split real bean yielded 2-3 times as much caviar (black dots) as the dried Amazon bean.

Here’s what the ice cream looked like when it was done and had hardened in the freezer:

photo (14)


I then organized a taste test panel comprised of 12 employees of the company where I work.  Their feedback provided another unexpected twist in the vanilla saga.  The averaged scores were (from high to low

  1. Amazon vanilla beans, dried to a safe water activity. Score = 7.5/10
  2. “Real” vanilla beans. Score = 7.3/10
  3. No vanilla at all (unflavored custard). Score = 6.7/10
  4. Vanilla extract. Score = 5.3/10
  5. Amazon vanilla beans, as packaged (wet). Score = 4.0/10

With only 12 participants, some of these scores are statistically indistinguishable from others.  But it’s fascinating that the dried Amazon beans have the highest score, while the Amazon beans when used straight from the bag have the lowest score.  Taste testers complained that the Amazon wet beans made the ice cream taste like chemicals, and left a lingering off-flavor.  I’m going to guess, though, that testers reacted to both what they saw and what they tasted.  Custard, made of egg yolks, cream, half and half, and sugar, tastes pretty darn good all by its self.  If you’re going to mess with that flavor, I think you have to improve it, which is what the real vanilla beans did.  In the case of Amazon vanilla beans, I think there was little impact on the flavor, but they gave the dish an appearance of having real vanilla beans.  That seems to be the winning ticket.  We had 2 testers with a lot of culinary experience, and those two both picked out the real vanilla beans ice cream as the clear winner.  I think the real vanilla packs the most powerful punch, too, but I knew which batch was which, and was probably biased.

Time for a brief summary.  I think the Amazon beans may be useful as a way to show you used real vanilla, but they must be dried, and have little flavor if you actually do that.  What’s more, they make your dessert taste like chemicals and create a mold hazard if they aren’t dried.  That’s not much of a bargain at any price.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Dec 12 / scott.campbell

Not So Magic Beans

A couple of weeks ago, I posted about a bag of vanilla beans I bought on Amazon.  These beans had an extremely high water activity when they arrived, something I’d been expecting since reading a few negative reviews on Amazon where customers complained of mold.  At a level of 0.98, I was pretty sure they would mold if left at room temperature in a zip lock bag (as instructed by the manufacturer).  I thought that drying the beans out would keep them from molding.  That  much turned out to be true.

The actual drying of the beans was a simple, if tedious, process. I set the beans out on a wire rack, where they sat for about a week.  Here’s how they looked on the last day:

photo (12)

The drying took quite a bit longer than I anticipated, and although I only got down to a water activity of 0.88, I assumed that the combination of water activity and alcohol the beans were soaked in would be enough to prevent mold.  Again, this has so far turned out to be true.  I returned the beans to a zip lock bag.  A couple of weeks have passed since then, and I haven’t seen any mold in my dried beans.

Here’s where things start to turn weird, and a little fishy.  I still had a key  question that hadn’t been answered: why were my Costco beans dry while these ones were wet?   I decided to research how vanilla beans are typically manufactured.  Turns out, this is a complex process that can be explained in about 9 minutes by a helpful lady on YouTube with an exotic latin accent. When it’s finally done, she claims that a dry vanilla bean will last for 50 years if stored in a sealed glass container.  I agree with her. The vanilla beans I bought at Costco lasted at least 7 years, and tasted as good last month as they did when I bought them.  So why weren’t my Amazon vanilla beans dry when I got them?

I grew even more suspicious when I made a batch of yogurt and used one of my now-dried vanilla beans to flavor it.  As usual, I halved the vanilla bean, scraped out the vanilla “caviar” (the little black dots), and whisked it into the milk.  I also threw the spent pods in to make sure they imparted as much flavor as possible.  Yesterday, I started eating the vanilla yogurt, and try as I might, I can’t taste any vanilla.  The little black dots do make for attractive window dressing, they just don’t seem to have flavored the yogurt.  This is a clear contrast to batches I made with Costco’s vanilla beans last month, and even a batch I made last week with the wet Amazon vanilla beans straight from the bag.  Seems all my vanilla taste had evaporated when I dried the beans so they wouldn’t mold.

This leaves us with three possibilities.  First, my taste buds are broken.  I can test this hypothesis by doing some taste testing – not something I’m opposed to.  Second, these vanilla beans are processed using a technique I’ve never heard of that results in a bag of beans soaked in alcohol.  I’m not an expert in vanilla bean processing, so this is certainly possible.  Third, these beans have been used to make vanilla extract, bagged, and then sold as virgin vanilla beans.  When used in their “wet” state, enough vanilla extract could still be in the pods and caviar to deliver some vanilla taste, and nobody’s the wiser.  I must admit, though, that even my “moldy” beans get 4.5 stars on Amazon.  Customers don’t seem to be disappointed, except for the mold issue.

I will now be forced to make multiple batches of vanilla-flavored products to test these possibilities.  Ice cream, creme brûlée, brownies, vanilla sugar – I didn’t ask for such a heavy burden, but someone’s got to get to the bottom of this issue.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Nov 27 / scott.campbell

Vanilla Bean Mold and Water Activity

Because I work at a company that makes water activity meters, I have overly sensitive antennae for mold problems that can be solved with water activity.  The concept is pretty simple: if a product’s water activity is below 0.7 aw units, it’s probably not going to mold.  If it’s below 0.6 aw units, it definitely won’t mold, no matter what.

I recently ran out of a supply of vanilla beans I bought at Costco a really long time ago.  Since Costco doesn’t stock them anymore, I tried, a somewhat useful source for obscure or exotic shelf-stable ingredients.  I happened on a 4 oz. bag of Madagascar vanilla beans with a relatively high rating.  For a bag that might contain 20 or so beans, $20 seemed a good deal.  Then I noticed that a number of recent reviewers complained that after using the first bean, the remainder promptly turned into a mold farm.  Ahh -food and mold.  Sounds like a water activity problem.

Here’s the package of beans:

photo 1 (2)

I took the beans in to work and measured their water activity by cutting about 1/4 of a bean into a sample cup, and putting it into a water activity meter.  Here’s the sample ready to measure:

photo 2 (2)

Here’s the result of the test:

photo 3 (2)

So there are a few interesting things here.  First, the beans themselves are described as “pure bourbon Madagascar vanilla beans”.  They appear wet in the package, and you can easily see that there’s some kind of liquid in the package along with the beans.  Alcohol does have anti-microbial properties, but I suspect a water activity of 0.966 is just too high.  Even if the liquid is bourbon, it still has more water than alcohol.  Note that the unopened beans won’t mold because there’s no oxygen in the bag.  Once opened, not only will mold spores be introduced, but oxygen, too. If I were to put these in a zip-lock bag on the shelf (as advised by the package) there would almost certainly be a mold problem.  Not sure if the manufacturer ships them this way because it is fixated on selling a certain weight of product, and sending out a “dry” bean would mean that for 4 oz. of beans, you’d need to add more beans.  I’m more interested in how many actual beans I get for the money, since that’s how they’re used in recipes: one at a time.

How should we solve the problem of the mold menace?  By doing the exact opposite of the directions on the package.  I’m going to leave the beans on a wire rack overnight, and see how much water evaporates off.  Incidentally, this is the same strategy used in the Zombie Cheeseburger post.  The easiest way to make sure something doesn’t mold is to leave it on the counter and let it dry out.  Once I’ve reached a water activity of 0.7, then I’ll put them in a zip-lock bag and use slowly over the course of many months.  Oh- and I’ll post here on how that strategy works out.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Aug 8 / scott.campbell

Salami, Soppressata, Peperone Done

To be frank, they were probably done some time ago, but I’m just getting around to posting now.  Here’s how each turned out:


And now a closer look at the salami:



Of the three, this one looked the best.  Definition was good, as well as the taste to match.  Next, the soppressata:



This also turned out nicely.  The tangy punch of the fermentation process makes this particularly enjoyable.  Not as much heat as I would have supposed.  Lastly, the peperone:


You may or may not be able to tell from this picture, but there is not nearly enough fat in this.  As a result, the taste is weak, even though it has plenty of spices.  The universal rule: fat = flavor.

I’m still watching the readings I get from the soil moisture sensors I stabbed the salami, ham, and cheese with:

8-8-13 data


The light green line on the bottom is the salami – it continues to drop as it has for some time.  The dark green line is the cheese – it too is losing water at a pretty steady rate, even with the coating of butter on the outside.  Puzzling, though, that the prosciutto isn’t losing any water at this point.  It’s been going for about 2 months, and I’ve seen no change in either sensor that is in it (the green line at the top of the graph, and the green line 2nd from the bottom).  Both lines are flat.  Perhaps it’s losing water so slowly that we haven’t seen any changes yet.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Aug 7 / scott.campbell

A Very Cheesy Blog Post

Cheddar isn’t the only cheese that we at Meter have made. (Yes, there was the provolone mishap but I don’t know if we’re going to count that.) We’ve also tried our hand at Parmesan and Havarti cheeses.

On July 1st I began a batch of what I hope will turn out as Parmesan in a matter of several months to a year. Like the farmhouse cheddar that Scott made, it is also a pressed cheese. Below is a picture of the just-pressed cheese (the wrinkly upper surface is from pressing a not-so-neat cheesecloth wrapping job):

Freshly pressed Parmesan

Freshly pressed Parmesan

After pressing the Parmesan it was dunked in a brine for awhile and then put in a “ripening box,” an environment that allows for the cheese to age.  The Parmesan sits on a needlepoint mat (I read online that it works like a cheese mat) inside a Rubbermaid plastic container, which has been placed inside a refrigerator set at 50 to 55 degrees Fahrenheit  and fluctuating between 60-70% humidity. The cheese has since gone about a shade darker of cream, and has accumulated a bit of mold but not too much.  The surface of the cheese is pretty firm, so whatever mold has grown on the cheese is easily wiped away with a piece of cheesecloth dipped in a 1-to-1 solution of salt and white vinegar. Yesterday I took a core sample of the Parmesan for testing, and it came back with a reading of 0.9202 aw. The sample was very firm and solid and did not crumble, and I’m hoping that in a few months it will be fit to be grated and used like any other Parmesan. While this Parmesan is drier than the other cheese I made (we’ll get to that in a moment), it is still fairly moist, and much moister than that poor provolone from last year that got stuck in a 10% humidity environment.

About a week later I made a second cheese: dill Havarti, a semi-soft cheese that would not (and did not) take long to age. It took me several hours to cook and cut the curds, drain them, then press them.


Heating milk


Cutting curds

Curds with dill about to be drained

Curds with dill about to be drained

About to press the curds

About to press the curds

Pressed Havarti, one day old

Pressed Havarti, one day old

Scott’s cheese book said the Havarti was ready to be eaten after being pressed, but to me it still seemed to be a spongy, curd-y mass that could definitely still be worked with. So per further instructions from the book, I plopped it in a heavily saturated brine overnight to try to draw some of the moisture out. After the brine, the cheese was set to air dry at room temperature for another day before I stuck it in a ripening box of its own.

Over the next few weeks in the month of July, I went to check on my cheese, flipping each one daily (only the first two weeks for the Parmesan, now it only gets flipped twice a week), and wiping off any mold growth with a cheesecloth dipped in the vinegar salt solution.

As time passed, I found myself wiping considerably more mold off the Havarti than I did the Parmesan. This is because while both were in the same types of ripening boxes and stuck in the same refrigerator, the Parmesan, a firm cheese, was formulated such that it was less wet than the Havarti, a semi-soft cheese.

Mold isn’t always an undesirable quality in cheese, but generally Havarti doesn’t have mold growing on it when you buy it from the store. The Havarti I made got a variety of interesting molds growing on it; there were brownish spots, blue-ish green spots, and sometimes a slight white film. Most of the more powdery molds were removable, but I was only able to scrub off some of the brownish spots that seemed embedded in the cheese’s rind. Speaking of the rind, as time passed the rind (or outer surface of the cheese) grew harder and became almost yellowed, the edges in particular. Recently I was also noticing some cracks developing in the rind, which I took as a sign that it was time that the cheese should be eaten.

On Monday of this week we finally had this tool called a cheese iron or cheese trier arrive in the mail. A cheese iron like this one is used by professional cheese makers to take a core sample of cheese without having to cut the cheese open and disrupt the aging process. Now that we had this tool, it was time to take samples…and do a taste test of the Havarti! This is what the Havarti looked like yesterday (Tuesday the 6th) this week:

Havarti...ready to eat!

Havarti…ready to eat!

In order to make a comparison, I went to the store and bought a hunk of commercially made dill Havarti.

Store bought on the left, homemade on the right

Store bought on the left, homemade on the right

I noted before testing water activity that the store bought version was much more spongy and moist than my firmer round of homemade cheese.  Then I took samples with the cheese trier.

Piercing the Havarti with the cheese iron

Here’s where it got interesting. Check out the difference in texture of the samples! The homemade turned out to be pretty crumbly, almost with the consistency of feta (I guess we should have tried to keep the humidity up in the fridge…). The store bought was just like your average Havarti, pretty spongy and soft.

Homemade cheese sample

Homemade cheese sample

Store bought cheese sample

Store bought cheese sample

Upon testing their water activity, there actually wasn’t too huge of a difference. The store bought Havarti gave a reading of 0.9640 aw, while the homemade Havarti gave a reading of 0.9568 aw. Not too big a difference it would seem, but it was just enough of a difference that the homemade cheese was of a crumbly texture while the store bought was much smoother. It’s interesting how only a little change in water activity can have such an influence on quality of a product, as demonstrated in this cheese test.

As mentioned before I also tested the Parmesan which currently has a water activity of 0.9202 aw (sorry, I forgot to take a picture of the sample). We’ll have to see how much that reading fluctuates while the Parmesan continues to age. I also took a reading on Scott’s cheddar using the cheese trier to take a sample. Here’s what the cheddar looked like:

Cheddar sample

Cheddar sample

It was pretty dry and crumbly, and if you can see the little white flecks…yep, that’s mold. The cheddar read in at the lowest water activity of all the samples, reading at 0.8606 aw. I’m not sure if that means we should wipe off the mold and eat the cheese, or try to up the humidity in the fridge and see what happens.

After testing, it was time to cut up my precious homemade Havarti (yes, you can become attached to cheese that you’ve made and babied along for a month). Here’s what the insides looked like:

DSC00786 DSC00787

Pretty isn’t it? (if you can call cheese pretty) As you can see in the below pictures, the cheese was pretty crumbly and brittle, and it was a little challenging to slice it with a cheese slicer without the slice breaking into pieces.

DSC00791 DSC00792

I tasted both myself, and while the homemade one wasn’t really what I’d call Havarti like you’d get from the store…it was still good! Both were tasty, but were completely different cheeses. It was a bit dry in comparison to the store bought but had a nice flavor with a tang that probably came from some of the outer remains of mold.

Here is a comparison of the homemade and store bought slices:

Homemade (left) vs. store bought (right) Havarti comparison

Homemade (left) vs. store bought (right) Havarti comparison

Can you see from this example and these pictures how testing water activity can make a HUGE difference in taste, texture, and overall quality of a product?

Then it was time for the taste test. Our new company chef suggested I serve the cheese with apple slices (the ones I used were Fuji), crackers (I used Keebler Club crackers), and a little drizzle of honey on top. Best. snack. ever.

Out of 17 participants, 7 preferred the homemade cheese! The majority preferred the store bought, but that was to be expected. I found from the comments made on their feedback sheets that really it was just a matter of preference. Some people like smooth, moist, mild cheeses, while other people like crumbly, slightly tangy cheeses, and this was reflected in their comments. While one person said the homemade had an “overall nice flavor,” another person said the homemade cheese was “less yummy.” In the end, it was just whatever kind of cheese people preferred. I feel the test was successful, especially since everyone was happy because they got to have a nice afternoon snack of crackers, cheese, and apples.

While I had intended to make something like store bought Havarti, I ended up with something not quite like the Havarti we’re all used to. I am glad though that my homemade cheese still ended up tasting quite nice. Since I ended up with a pleasant tasting product despite the odds, I feel this first completed cheese venture was a success. Join us in forthcoming blog posts to see the progress of my Parmesan and Scott’s farmhouse cheddar, using water activity to understand how water activity readings relate to product properties.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Jul 22 / scott.campbell

Blog Turns 100 – Curing Meat Update

It’s been a couple of weeks since the last post – ok, maybe longer.  Still, this is the 100th post on, and to celebrate that fact let me just say that most people have an irrational affinity for round numbers.  Since last post, the cured meats have continued to cure, along with our farmhouse cheddar.  Things look pretty good for both, so let’s check in on the moisture and salt numbers as measured by our spiky probes.

The cheese didn’t appear to be losing any water for quite a while, but slowly began to dry out about a month into its aging process.  There is a really strange data spike in the middle.  This is where I took the probe out to address a mold issue on the chesse’s surface.  After I re-installed the probe, it spiked to a much lower reading, gradually increased, and then started to decline on a path similar to where it had been before removing the probe.  Strange, but the upshot is that the cheese started at 1.01 meter cubed/meter cubed, and now reads .86.

The salami has declined steadily to the point where it reads 0.45 meters cubed/meters cubed. This is compared to its starting value of 0.59.  Roughly, it’s lost 1/4 of its water by that measure.  I have been taking periodic water activity readings to compare, and here’s what I found:


Salami data - 7-22-13


This isotherm suggests a simple calibration curve that is specific to the salami that can tell us when it’s done.  If I were just looking at the data, I would have said the salami was “done” at a water activity of around .780, which corresponds to a water content reading from the probe of 0.44 meters cubed/meters cubed.  So that’s just for the Tuscan Salami, but at least if I make it again, I won’t even need water activity readings – just water content probe readings will suffice.

The Prosciutto doesn’t seem to be doing much of anything.   Luckily, it’s not molding, but it doesn’t seem to be losing a lot of water either.  Here are the readings:

Ham data 7-22-13


The green lines are water content, and the brown one is electrical conductivity.  There really hasn’t been a shift yet.  I don’t think that worries me, as it still has over 10 months to get things going.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Jun 19 / scott.campbell

Sensors Signal Safe Setup

It’s been almost three weeks since I loaded a bunch of new meats into the fridge and started taking remote readings on their curing process.  I’m happy to report that I have a new set of squiggly lines to share.  Recall that one of the things we were trying to measure was how salty the meat became during the salting process.  Now that I’ve pulled the ham from its salty hibernation and have hung it to dry, I’m interested to know whether the saltiness has changed.  That brings us to the squiggly lines:

Ham Data 6-19

The bold brown line is the “saltiness” measurement – electrical conductivity.  What are we seeing?  Sometime on June 10th, I pulled the sensors out of the ham, rinsed off the salt it had been sitting in, packed it with lard, peppercorns, and some red peppers, and hung it in the fridge to cure.  This shows up as a massive negative spike on the graph – useful because I don’t need to remember exactly when this happened.  What we see is that, up to the change, more and more salt was being taken up by the ham.  After removing it, however, there has been essentially no change to its salt content.  The ups and downs on the graph are likely due to changes in temperature, and overall, readings haven’t changed much.

That temperature point is also important to consider.  Every time we see temperature go up, we see moisture content reading increase, and the salt reading decrease.  Both of these changes are likely unrelated to actual trends, but the sensors appears to show a small amount temperature dependence.

The large cycles in humidity that you see on the graph result from trying to get the fridge humidity down in the 60-70% range.  I do this by leaving the fridge door open for a day or two.  As yet, the lowest I’ve achieved with the door closed has been 80% – still too high if we want to avoid mold growth.     Water activity theory tells us that most mold won’t grow if humidity/water activity is below 70%/0.7 aw units.  Each time I shut the door, the still-kinda-new sausage is releasing enough moisture that it raises the humidity into the danger zone.  It’ll eventually be fine, and I’ll be back to using the humidifier to maintain a 60% humidity level.  The nice thing about the remote sensor is that I check the humidity from my desk, and then run out to the fridge if the reading is too high.

So what of the moisture readings?  The ham doesn’t look like it has lost any water at all – the green line is flat.  The salami, however, has declined from 59% moisture to 56% in about a week (the light green line is the water content):

Salami Data 6-19

We expect the salami to be done in a couple of weeks, while the prosciutto will take almost a year, so no surprises here yet.  In the next post,we’ll look at the data from our cheese that’s been aging for a couple of weeks now.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Jun 10 / scott.campbell

Prosciutto Packed


Any time a charcuterie experiment doesn’t go horribly wrong, I’m pretty much satisfied.  So I was happy when I dug the prosciutto out of its mountain of salt and found nothing was amiss.  It had lost some water into the salt, but not enough to actually have liquid present.  Also the flesh and skin both seemed to have absorbed the salt well, but the ham was still squishy.  The exposed meat had taken on some of the translucence of heavily salted protein.   In this picture, I’ve already coated the ham with lard and pepper, and am about to wrap it in cheesecloth per the instructions.

This all brings me to the data – how much “saltier” is our ham than when we started?  The first reading we registered for electrical conductivity was maybe 30 minutes after we put the probe into the ham.  We got a reading of 0.231 mS/cm (don’t ask me about these units).  The last reading we got was 2.49 mS/cm – roughly 10x higher.  If we had left the meat in the salt for a day per pound (as Ruhlman suggests), I think we would have had 10-15% more salt in the meat.  Not sure if that will make a difference but we’re taking the ham out of the salt about a week earlier than the recipe suggests.

Here’s what our probes are telling us:


The brown line is the “saltiness”, and you can see that while it increased the entire time the probe was in the ham, the rate at which the salt went into the ham was steady, and then began to decrease on day 10-11.  This may be because the probe doesn’t extend all the way to the bone, or it could just be that the ham wasn’t taking on much salt at that point.  You can also see the green line saying that the moisture content of the ham is increasing.  It wasn’t.  Rather the massive salt invasion into the ham was affecting the moisture readings.  Since the ham was actually losing water during this time, we know the actual moisture content was decreasing.  As the salt content began to stabilize, so did the moisture content readings, and there was even a small decline recorded, something we anticipate seeing more of during the full year the prosciutto will be curing.  Here’s how it looks all tucked in for a year long slumber (with probes re-installed):









Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>

Jun 3 / scott.campbell

Mold Flourishes at High Water Activity

As if hastily applied by unseen nocturnal mold gnomes, a lush white layer of chalky mold appeared on the tuscan salami by the time I checked it yesterday (see picture).  This was mostly


by design.  I sprayed the surface of these sausages with mold starter, and thought that keeping the fridge humidity (or water activity) really high would give the mold spores the environment they needed to do their thing. The humidity/water activity in the fridge appears to be 95%/0.95aw.  This is by far the most successful mold hatching I’ve had, so I’m probably a little too excited about it.  We had bad mold (black, green, fuzzy) show up on the first sausage-making attempt.  In retrospect, conditions were no different than they are right now in the fridge, but we hadn’t put mold starter on the surface of the sausages back then.  Our successful mold growing shows that the trick is getting the good mold to out-compete the bad mold, rather than hold a fridge water activity so low that nothing can grow (probably below 0.7 water activity, or 70% humidity).

On Friday, I also made up two more batches of sausage: soppressata, and peperone. These I stuffed into dried hog casing, which made me wonder why I bother with the gross and rather rank hog casings.  They even look better, if a bit “standardized”.  DSC00657


The soppressata is to the left, with the darker peperone to the right.  You can see the stick in the foreground that has the soil moisture sensor in it – I’ll be taking dry-down readings on this too, as well as readings on the moldy tuscan salami (you can see the smaller black cable of another sensor sticking out of one of the sausages in the pic at the top of this post).  The prosciutto appears to be coming along nicely – more on that later.

Download The Food Manufacturer’s Complete Guide to Water Activity—>

Download The Food Manufacturer’s Complete Guide to Simplified Shelf-Life Testing—>