1

Water is the cheapest ingredient available. It doesn’t cost much to add water. Unless you

add the wrong amount. Then water can be incrediblyexpensive. Add too much and your product will moldon the shelf. Get the amount wrong, and your productwill lose its crunchiness or softness. Combine twoingredients with incompatible water levels and get aclumpy mess that gums up the factory machinery. Storea combination of ingredients at unstable moistures and endup with an ingredient so hard and dried up that it breaks acustomer’s tooth and causes a lawsuit. For a cheapingredient, water can cause a lot of expensive headaches.

How to profit from proven product-qualitytechniques in 5 quick lessons

more about water and your product inside D

DECAGON

Water activity tools

2

Lots of food processors try toget a handle on the problem byfinding out how much water isin the product. They weigh, dry,weigh, burn up a lot of product, andin the end get a crude measurementthat doesn’t answer their questions.

Quantity doesn’t tell you muchabout quality. And good foodquality is what good food science isall about. Measure water activity,and you have a fast, accurate,powerful measurement of foodquality. Know the water activity,and you know not just how muchwater is in your product, but whatthat water is doing and what it willbe doing as the product sits on theshelf. Water activity lets you seethe total quality picture. Andwhen you see the whole picture,water once again becomes thecheapest ingredient in yourproduct.

CONTENTS

Dollars andcents

1: Drying healthy

dog food

page 3

Fresh bakerytransport

2: Grocery

distribution

& shelf-life

page 4

A soup mix story3: Caking &

clumping

page 6

Isothermcomparisons

4: Pecans & Airbags

page 8

moisturemigration

5: Fruitcake & Fruit

cereal surprise

page 10

3

Dollars and cents#1: Drying healthy dog food

PRODUCT SAFETYand quality is the bottomline. Spoilage and poorquality hurt or destroysales. You set qualitystandards and stay withinthem. But where do youset the standard, and howmuch of a safety margin doyou need? Too manymanufacturers measuremoisture content and use aby-guess-and-by-gollystandard. For example, apet food manufacturer weknow was producing to8% moisture content.Why? Because at that level,he had never had spoilageproblems or shelf lifeconcerns. He had a safeproduct. But he discoveredthat he was drying upprofit. Then he used wateractivity measurements toget a more accurate foodquality picture. The wateractivity of his product was0.50 aw. A little additionalinformation revealed thathe needed only to staybelow 0.65 aw to produce aquality shelf stable productthat was below the moldgrowth limit.

Without knowing his exactcosts, we can still infer theprofit margin representedby a difference in 0.50 awand 0.65 aw. If he produced20,000 pounds of productper hour, operating sixteenhours a day five days aweek, and sold the productfor forty cents a pound,

raising the target aw

level would allowhim to producean additional1,877,314 poundsof product per yearand would generatean additional$750,925 in revenue.Even better, much of thatwould be profit since thischange would reduceutility costs while keepingman hours, raw materialcosts, and wear and tear onequipment the same.

Just a tricky little storyproblem? Not at all. Manymanufacturers havedecreased costs andincreased profits usingwater activity. Moisturecontent measurements justcan’t give you the safetyand quality informationyou need. Knowing watercontent reveals nothingabout the total food qualitypicture. Good operatingdecisions can only be madeusing good information.Usually, water activity isone of the best pieces ofinformation you can havein setting and maintainingfood quality standards.

The food industry, togetherwith the FDA and USDA,is adopting a system toimprove safety and reducethe incidence of food-borneillness. The HazardAnalysis Critical ControlPoint program (HACCP)

involves settingcritical controlpoints thatmeasure foodqualitythroughoutproduction.

Rather than“check-and-chuck” at theend, these critical controlpoints detect problems allalong the production line.Water activity is a criticalcontrol point in manyprocesses. It defines criticallimits for microbial andchemical parameters whichmust be controlled toprevent food safetyhazards.

Even when safety isn’t anissue, water activity is animportant measure ofquality. You can set criticalcontrol points to avoid lossof crispness in dryproducts, caking andclumping of powders,tough and chewy texturesin moist products andshortened shelf life. Wateractivity is a powerfulnumber. Measure andmanage it, and you gaincontrol over the quality ofyour product. You canproduce a consistentlyhigh-quality product andmeet expectations withouthaving to throw awaymistakes—and you won’tneed a huge safety marginthat wastes potential profit.�

4

Fresh bakery transport#2: Grocery distribution & shelf-life

MOM AND POPBakery baked exceptionalcakes. Local demand washigh. Mom and Popwanted to increase sales byexpanding. Unfortunately,their delicious cakes had ashelf life of three days. Athree-day shelf life left notime to ship the cakes.How could they keep their

quality andsimultaneously

get longershelf life?Once

again,water

activity is theanswer. Water activity is a

well known predictor ofmicrobial growth. A simplechart shows water activitiesbelow which certain moldswill not grow. And there’sjust one chart. These valuesare the same whether youare measuring the wateractivity of brie or dogfood. They don’t changewhen you change yourrecipe.

Why is water activity thebetter predictor? Watercontent measures quantity,it tells you how muchwater is in the product.Water activity measuresquality because it tells youwhat the water is doinginside the product. Somewater is ‘bound’—tied upchemically within theproduct. The rest is ‘free’.The ‘free’ water may beused by passingmicrobes—but only if thereis enough available. Somemicroorganisms need morewater while others needless, but all need a certainamount of ‘free’ water.Moisture content tells onlyhow much water there is,not how much is available.If you measure the wateractivity of your product,you will know whichbacteria, molds, or fungican grow on and in it. Byreducing the water activity,you can rule out thegrowth of whole classes ofmicrobes. At low wateractivities (>0.60), you canpreclude the growth ofanything at all.

As Kimberlee J. Burringtonwrites in Food ProductDesign (Aug. 1998),“Development of manybaked products involvesmaximizing moisturecontent to produce the best

possible eating qualitieswhile minimizing aw.Lowering the aw increasesproduct stability in termsof susceptibility tomicrobial growth.”

You can reduce yourproduct’s aw in severalways. Adding humectantslike salt and sugar ‘binds’the water in the product.You haven’t reduced thewater content, but youhave reduced the wateractivity. You can dry theproduct in a number ofways; air drying, vacuumdrying, freeze drying, andosmotic dehydration allreduce water activity. Youcan also lower wateractivity by lowering thetemperature of theproduct. In many cases, aminor recipe change canbring water activity to asafe and shelf-stable level.

Mom and Pop solved theirproblem with a minorchange. They manipulatedtheir recipe while keepingan eye on the changingwater activity. The finalproduct kept all thetastiness of the original butwith a lower water activity.Their new 9–12 day shelflife made shipping—andsuccessful expansion—possible. �

5

� Shelf-life Tale“ … I don’t think [Hostess has] made a Hostess Twinkie in 45years. Think of the ef f iciency. Churn out 800 million Twinkies ever y50 years, then f ire ever ybody but your marketing and distributionpeople.”http://www.textf iles.com/magazines/M00SE/m00se21

� Twinkie TruthsRecommended Twinkie shelf-life—14 days.Twinkies baked each year ≈ 500 million.http://weeklywire.com/ww/04-10-00/chicago_a.html

Highly perishable (fresh) foods andcanned fruits, vegetables, meat,f ish, and milk

Some cheeses (Cheddar, Swiss,Muenster, Provolone), cured meat(ham)

Fermented sausage (salami),sponge cakes, dr y cheeses,margarine

Most fruit juice concentrates,sweetened condensed milk, syrups

Jam, marmalade, marzipan, glacéfruits

Jelly, molasses, raw cane sugar,some dried fruits, nuts

Dried fruits containing 15-20%moisture; some tof fees andcaramels; honey

Dr y past, spices

Whole egg powder

Cookies, crackers, bread crusts

Whole milk powder; driedvegetables

1.00–0.95

0.95–0.91

0.91–0.87

0.87–0.80

0.80–0.75

0.75–0.65

0.65–0.60

0.60–0.50

0.50–0.40

0.40–0.30

0.30–0.20

Pseudomonas, Escherlchia, Proteus, Shigells,Klebsiella, Bacillus, Clostridium perfringens, someyeasts

Salmonella, Vibrio parahaemolyticus, C.botulinum, Serratia, Lactobacillus, Pediococcus,some molds, yeasts (Rhodotorula, Pichia)

Many yeasts (Candida, Torulopsis, Hansenula),Micrococcus

Most molds (mycotoxigenic penicillia),Staphyloccocus aureus, most Saccharomyces(bailii) spp., Debaryomyces

Most halophilic bacteria, mycotoxigenic aspergilli

Xerophilic molds (Aspergillus chevalieri, A.candidus, Wallemia sebi), Saccharomycesbisporus

Osmophilic yeasts (Saccharomyces rouxii), fewmolds (Aspergillus echinulatus, Monascusbisporus)

No microbial proliferation

No microbial proliferation

No microbial proliferation

No microbial proliferation

Range of awMicrooranisms Generally Inhibited by

Lowest aw in This RangeFoods Generally within

This Range

* Adapted from L.R. Beuchat, Cereal Foods World, 26:345 (1981).

Water Activity and Growth ofMicroorganisms in Food*

6

IN HIGH, dry SaltLake City, a salt shakeris a salt shaker. Turn itupside down, and thelittle grains pour out.In rainy Seattle, whenyou flip and shake,you are much morelikely to hear a“thunk” and get nextto nothing. Evenaround the house,clumping and cakingcan cause problems.On the production line,those problems get big andexpensive.

A small dry soupmanufacturer wasdetermined to avoidclumped-up soup base.One day he was processinga base with 3% moisturecontent. He measuredsome newly receivedpepper and found it to alsohave 3% moisture content.Sure that it was safe to mixthe two components, hedid so and headed out tolunch. The manufacturerreturned from lunch,flipped on his processing

machinery anddiscovered thatthe whole batchhad clumped.He had tothrow awaythe lot. Hisloss felt evenworse as hespentconsiderabletime cleaningup his

equipment.

Caking and clumping areubiquitous problems in thefood and pharmaceuticalindustries. Predicting—andavoiding—the problemmeans paying attention tothree things: time,temperature, and wateractivity. Over time, freeflowing powders will movethrough the stages ofcaking. The particleswithin the powder will getwet, then sticky. Stickyparticles will agglomerate,compact, and finally reachthe liquefaction stage. Theprocess is affected byparticle shape and size,

WATER ACTIVIT Y does not per fectly

describe the chemical

processes occurring during

clumping and caking. Some

products, like milk powders,

may contain amorphous

cr ystalline compounds. These

products are subject to a

phenomenon known as “glass

transition”. When processed or

stored above the glass

transition temperature, these

powders show stickiness,

collapse, and cr ystallization

due to the viscosity decrease

above Tg, and the increased

plasticization by water. During

collapse, moisture is released

from the crystallized to the

amorphous regions, leading to

the formation of inter-par ticle

liquid bridges, causing

par ticles to stick together.

There is no routine method

currently available to monitor

glass transition during

processing. However, by

monitoring water activity and

keeping the product below a

specif ic water activity, caking

and clumping can be avoided

even in glassy products. �

aw

GL ASSY PRODUCTS

� Generally, air drying of aromatic herbs is an ef fective method ofpreser vation that inhibits the growth of microorganisms and delaysthe onset of some biochemical reactions in the f inal product.—From Yousif,A.N. (2000) J. Food Sci. 65:926-930

7

A soup mix story#3: Caking & clumping

temperature, moistureavailable within the system,applied pressure, andchemical composition. Theproblem can be avoided bykeeping the powder belowa certain water activity.

The soup manufacturerthought he had the processunder control. His soupbase was dry, thetemperature and relativehumidity in his processingplant were fine, so why didhe end up with a mess?This manufacturer wasdistracted by moisturecontent numbers. Just likethe fruitcake producer inLesson Five, he thought hehad his eye on the ball, buthe was actually out of thegame completely. Hethought that dry enoughfor soup base was also dryenough for pepper. But itisn’t the quantity of water,it’s the quality—not howmuch there is, but howmuch is available. And the3% water in the pepperwas free enough to wreakhavoc. Had he measured

Solutions to preventcaking of powders:

� Dry powder to lower aw.

� Treat powder at low humidity atmospheres.

� Package powder in high moisture barrier packaging.

� Use in-package desiccation.

� Agglomeration

� Add anti-caking agents.

water activity, he wouldhave found that the soupbase had a water activity of0.28, while the freshlyground pepper was at 0.69aw. When he combined thetwo, he brought the wateractivity of the mixtureabove the critical clumpingpoint.

The critical point at whicha product will cake orclump is product-specific.It must be determined byputting the product atdifferent humidities andmeasuring the wateractivity at which theproduct begins to clump.Maintain the productbelow that value, and youwill avoid clumping. As thesoup manufacturerdiscovered, addingingredients at higher wateractivities can change thewater activity of theproduct. In addition, highambient humidity andincreases in temperaturecan raise the water activityof the product. �

PREVENT CAKING

� The total number of outbreaks attributed to contaminatedspices in the last 30 years is low and no repor ted outbreak hasoccurred in the United States. Since several bacterial pathogensand pathogen indicators may at times be isolated from herbs andspices, such a low number of outbreaks may inpar t be attributed to the role of spices asingredients, and the inhibitory or lethalef fects of low water activity. —From theAmerican Spice Trade Association (ASTA)

8

Isotherm comparisons#4: Pecans & Airbags

All the varying watercontents are related to thesame water activity. Asimple chart shows thewater activity values belowwhich certain molds do nothave enough ‘free’ water togrow. For example, below0.6 aw, water is so tightly‘bound’ that it isunavailable to even themost xerophytic fungi.That number is the samefor every product,ingredient, and substance.

Water contentmeasurements are crudebecause measurementmethods may not be veryaccurate. A pecan growerdiscovered this in thefourteenth year of hisoperation. He grew pecansfor school fundraising.Every year he dried hispecans to a moisturecontent of 4%. Forthirteen years, the method

worked perfectly. Yearfourteen could have

been titled “TheReturn of the MoldyPecan.” He wasbewildered.Understanding wateractivity gave the totalfood quality picture. A4% moisture content

relates, in pecans, to0.68 aw. That’s

pretty safe.Anythingunder 0.60 aw

will be totally free frommicrobial growth. In

pecans, that relates to3.8% moisture content.Unfortunately, themoisture contentmeasurement was onlyaccurate to ±0.5%. Thatmeans his fourteenth yearpecan shipment hadprobably closer to 4.5%moisture content and wasoutside the safe zone.

Using an aw meter solvesthe accuracy problem. Thedifference between thepecan producer’sborderline numbers— 0.68and 0.70— is huge to awater activity meter withan accuracy of ±0.003 aw.The pecan manufacturercan measure water activityand be confident ofproduct safety with everyshipment.

In fact, the better accuracy,speed and easymeasurements of wateractivity are attractive evento people who need tomeasure water content. Anairbag manufacturerneeded to control watercontent in the chemicalpropellant (cupric oxide)used in the airbags. Heneeded a tight control. Ifthe cupric oxide’s watercontent wasn’t within spec,the airbag wouldn’t inflate.Cupric oxide has a watercontent of about 0.05%—so low that measurementswere hard to make. Theyrequired a high precisionmoisture balance and

� Nothing is morememorable than fish f lakesroasting in a water content

drying pan.

MANY manufacturersuse the “loss-on-drying”method to measure watercontent. They weigh theoriginal product, dry it out,and weigh again. Since“dry” means really dry, andsince it usually needs to bedone in a hurry, manypeople rapidly dry theproduct in a vacuum oven,microwave or moisturebalance for that middlestep. As one fish processornoted, “This is a reallystinky process.”Unfortunately, it’s alsoslow, crude, and oftenmisleading.

Misleading because whenyou measure moisturecontent, you need to knowhow dry is dry enough.The water content of a safeproduct varies fromproduct to product andfrom recipe to recipe. Incorn seeds, safe storagemeans a water content of12.5% while inmustard seeds,it’s 7.5%.

8

aw

9

� During the three seasons, 1999-2001, the utilized production ofpecans averaged 310,650,000 pounds per year and grossedgrowers $261,846,000.

Pecan meats sold as halves are premium-priced products used fordecoration. Broken nut meats are used in baked goods,confections, and ice cream.

About 78 pecans go into a pecan pie.

scrupulousattention toscientifictechnique. Evenunder the bestconditions,readings werehard to get.After findingthe relationshipof water contentto water activityfor cupric oxide,they found thata change assmall as 0.02%water contentrelated to achange of 0.2aw. The aw meteraccuracy of

±0.003 aw let themmeasure to tightercontrols—and made themeasurements much easierto get. �

But I really really really need tomeasure moisture content!Water activity is always themost relevant and power fulmeasurement of food quality.But some manufacturers alsoneed to know moisture content.They may be selling on a weightbasis, have labeling issues, orneed to determine cer tainnutritional values. If thisapplies to you, it is possible tohave the best of both worlds.Water activity and watercontent are related. Thisrelationship, called themoisture sorption isotherm, isunique to each product. Itchanges depending on whetherit was obtained by wetting ordr ying the product. Typically,large safety margins are builtinto water contentspecif ications to allow for theseuncer tainties. A vegetabledehydrator needed to measurewater content. They wanted thespeed and accuracy of

measuring water activity on theproduction f loor. Theydeveloped a sorption isothermfor their product (see table).They needed to dr y to a 6%moisture content specif ication.The 6% moisture correspondsto 0.462 aw using their uniqueisotherm cur ve. Once theyswitched to measuring wateractivity, they discovered thatthey had much better controlover moisture content. Their oldmoisture balance method, inaddition to being slow andpainstaking, was only accurateto ±0.5%. That meant productmoisture content varied from5.5% to 6.5%. The ±0.003 aw

accuracy of their water activitymeter gave them much greaterprecision. They were able tospecify a ±0.01 aw tolerance onthe production line. Nowmoisture content in theirproduct varies from only 5.86to 6.14%.

A sorption isotherm is ideallyproduced using the processthat will bring the product to itsf inal water content—baking forcakes, standard drying processfor potato flakes. To getspecif ic information ondeveloping the isotherm,contact Decagon. �

k AirbagDeployment

DEVELOPING AN ISOTHERM

Moisture Sorption Isotherm201816141210

86420

0 0 .1 0 . 2 0 . 3 0 . 4 0 . 5 0 . 6 0 . 7 0 . 8 0 . 9 1

Moi

stur

e C

onte

nt

Water Activity

10

MixedCandiedFruit0.862 aw

Dough0.857 aw

moisture migration#5: Fruitcake & Fruit cereal surprise

EVERY FOODproducer needs to know

what will happen totheir product as

it sits ontheshelf.Shelfstable

can mean“won’t get

moldy,” butit covers a lot

of other foodqualities as well.

Take, for example, awell known raisin

bran manufacturer. Boththe manufacturer and the

consumer of the raisin branexpected crunchy flakesand chewy raisins. Theconsumer got a surprisewhen he bit into a rockhard raisin and broke atooth. The manufacturersoon after got a surprise inthe form of a lawsuit.Fortunately, not allmoisture migrationproblems end up in a courtof law. But if you produceand sell products thatcontain discreteingredients, you need toknow where the moisturewill go as the product sitson a shelf.

A fruitcake manufacturerwants to predict conditionsover time in her fruitcake.She doesn’t want the caketo get soggier as the fruitpieces get dry and hard. So

she measures moisturecontent. The cake contains30% water, while the fruitpieces have 50% water.She knows that waterwants to come toequilibrium. Therefore, sheassumes that water willmigrate from the wetteringredient (the fruit) intothe drier ingredient(the cake).

Unfortunately, she has hereye on the wrong ball.While she’s watching thewater content and feelingconfident of the score, awhole different game isgoing on at the wateractivity level. If shecontinues without knowingthe score of that game,she’ll suddenly beconfronted with a surpriseending—in this case, driedup cake and soggy fruit,because the cake had ahigher water activity thanthe fruit pieces at thesemoisture contents.

In this case, water contentis nothing more than adistraction. Pay attentionto that number, and the

outcome feels like sleight-of-hand. True, unless actedon by other forces, waterdoes want to come toequilibrium. But,equilibrium occurs whenthe partial specific Gibbsfree energy is the sameeverywhere in a system.Water activity is ameasurement of Gibbs freeenergy. Water content hasnothing to do with theenergy of the water. Tounderstand this conceptbetter, think of two tanksof water. One is almost fullat 10,000 gallons. Theother huge tank is nearlyempty, containing only onegallon. Which way willwater move? Knowing thewater content of the tanksis completely misleading.The volume of water isirrelevant. Water movesfrom higher pressure tolower pressure, not fromfull to empty. If we raisethe pressure of the nearlyempty tank by putting iton top of the almost fulltank, that last gallon ofwater will quickly leave itsspacious tank.

11

Likewise, water activity,not water content, predictshow water will migratewithin a product. Thefruitcake manufacturer candevelop a recipe in whichcake and fruit pieces bothhave the same wateractivity. No moisturesurprises when thiscake is stored andsold—it is a safe,palatable, andshelf-stableproduct.

Knowing the wateractivity of discreteingredients can help youdecide how to processthem. One solution to amoisture migrationproblem is to lower orraise the water activity ofdiscrete components untilthey have the same water

“Sure enough, the next Christmas … mymother, installed [the fruitcake] as a

doorstop …someone ate a few slicesof this loaf, [it] is still as pristine as

it was over 10 years ago!”—The Fruitcake Hate Page

www.anet.net/~penfold/ihatefruitcake.html

Dough 0.857 24.5

Fruits (mixed) 0.862 52.2

W

TheLessonChart

activity value. You can alsoretard the diffusion processwithin a component byincreasing its viscosity. Anedible barrier, like

chocolate

coating on theinside of an ice cream cone,can prevent moisturemigration. Sometimes awdifferences that can’t beequalized require separatepackaging. �

� Determines direction of

moisture transfer

� Most degradative reaction

rates increase with

increasing water activity

� Most degradative rates

correlate better with water

activity than moisture content.

� Undesirable textural changes

due to moisture transfer

� Crispness, crunchiness lost

if aw > 0.4

� Sof tness, chewiness lost

if aw < 0.5 - 0.6

WATER ACTIVIT Y

12

DECAGON950 NE Nelson CourtPullman, Washington 99163

800-755-2751fax 509-332-5158aqualab@decagon.comwww.decagon.com/aqualab

AquaLab

Water Activity Tools

5quick

lessonsAquaLab

©2003 DECAGONPrinted in USA

X AquaLab's

chilled-mirror dewpoint

technology gives you

superior speed and accuracy

for water activity

measurement.

AOACAPPROVEDMETHOD