IAN PAVEY

Principal Electrostatic SpecialistChilworth Technology P. Ltd, 0-12, Rishabh Complex, Opp.To Fun Republic

New Link Road, Andheri (W), Mumbai-400 053 Tel: 022-66942350 - 51

Introduction

There is no doubt that static electricity costs thepharmaceutical industry dearly. The cost can often bemeasured in terms of production rates, yields, and downtime in a wide range of operations. Unfortunately whenstatic leads to damage to plant, environmental damage,injury to personnel, or even foss of life, the cost, in humanterms at least, may be quite immeasurable.

r-- Hardly anyone in the process industries is unaware. of static electricity and some of the problems it causes.

Indeed, electrostatic phenomena feature in some of theearliest recorded scientific observations. And yet, 4500years later the feeling that it is unpredictable and dealingwith it is a "black art" is stili a commonplacemisconception, At the same time, electrostatics is playingan increasingly important role; across the whole gamutof process industries - and none more so thanpharmaceuticals. As we shall see, in many respects, thepharmaceutical industry is particularly vulnerable:

',

However, the fact is that electrostatics is now better

undrrstood than ever before. This article will explain

briety where the charge comes from and how it leads toa n mber of different types of problem with referencesto s ecific operations where static has proved to be ahaz rd. With understanding comes logical solutions sothe article also explains the basics of a systematicapp oach to dealing with the problems described, whichcan often readily be extended to other situations oncethe fundamentals are properly understood.

Le~islation, Standards and Guidelines

I egislationin the general area of process safety hasofte appeared rather vague. It is clear that there is anobli ation on the part of employer.s to run an operation,whi h does not put their staff (or the public) at risk.Unf rtunately national legislation has often fallen shortof i dicating exactly and unambiguously how this shouldbe one.

Irlowever, over the next few years European Memberstat~s will be adopting a new and far reaching directivein tris area (ATEX 137) which leaves little room fordisqussion or misinterpretation.

Employers will be obliged to identify and classify areaswhere flammable atmospheres could occur, includingdust handling areas. Also, strategies to cont 01and avoidpotential ignition sources must be devise Inevitablythis will require the detailed kno'wledge of the fl mmabilitycharacteristics of all materials, includin ignitionsensitivity. In addition, plant will have to be d signed tosafely control the consequences of an explo ion. Thiswill require a knowledge of the severity of any explosionas a result of materials being. At last the need to properlyknow your process materials, wh'ich safety experts agreeis crucial, will be enshrined in law.

For many years British Standard BS5958 andGermany's ZH1/200 have stood out as general guidancefor industry on the avoidance of hazards due to static'electricity. Recently a CENELEC Technical Report(R044-001 :1999) has been published. Entitled "Safetyof machinery: Guidance and recommendations for theavoidance of hazards due to static electricity" it discussesa wide range of processes and situations. It is c~rtainlya great help in standardising the advice given by thosealready expert in the area although realistically, workingfrom a document such as this with no previousexperience may be a little like learning to drive withreference to the Driver's Manual which comes with a

new car! In theory it should be possible but the processis fraught with risks along the way. Indeed, this latestdocument still has to use the phrase "expert adviceshould be sought" within its pages.

Static Charge Generation

The two most important ways in which unwantedelectrostatic charg(3 is acquired are induction chargingand tribo-charging. It is crucial that the principles of eachare understood in order to recognise where charge maybe produced, which is an essential precursor to dealingwith it.

Tribo-C.harging

Whenever two different materials c;ontact one anotherelectrons will move across the interface so that one

becomes negatively charged (excess electrons) and theother positively charged. If the two materials are goodconductors (such as metals) all the exchanged charge

C.1emical News, January 2009 35

will flow back through the last point of contact when theyare separated. However, if at least one of the materialsis a poor conductor this will not happen and the chargewhich was exchanged during the contact will be carriedaway on separation.

It is extremely important at this stage to dispel onevery common misunderstanding. Only one of the twocontacting materials must be a poor conductor for bothto carry charge away on separation. Furthermore, if thegood conductor (a metal, perhaps) is well earthed chargewill still cross the interface and the poor conductor willstill carry away charge when they separate. The onlydifference is that the charge the metal acquired will belost to earth almost instantly. All too often it is thoughtearthing plant solves all electrostatic problems. Thereality is that although earthing plant is vital, it is not thewhole answer.

The magnitude of charge acquired by tribo-chargingdepends on various factors but in general the moreenergetic the process the greater will be the chargegenerated. Table-1 illustrates this by giving typical chargelevels acquired by powders undergoing some verypommon processes.

Table 1: Typical Powder Charge FollowingCommon Processes

Charge:Mass Ratio (IJC/kg)10-3 to 10-5

10-1 to 10-3

1 to 10-2

1 to 10-1

102 to 10-1

103 to 10-1

rocess-ievingouringcroll Feed Transfer

rindingicronisingneumatic Transfer

nduction Charging

All but the most insulating of materials will charge bynduction. When exposed to an electric field (such ashen in the vicinity of a charged object) opposite chargesithin the material will tend to separate - either beingttracted towards, or repelled from, the nearby charge.,ny local excess of charge at a point of contact will theneak away according to just how conducting the materials and how good is the contact. This will leave behind anverall excess of the opposite sign of charge.

In order to clarify this point a typical example of howhis may occur in practice is shown in Figure 1. Figure(a) illustrates a person (a very good conductor) near to

6 Chemical News, January 2009

a big bag (FISC) containing a highly charged product(plastic pellets, perhaps). The separated charges areshown, as is the negative charge leaking away via shoesand floor. Figure 1(b) shows the person now moved awayfrom the vicinity of the FISC, carrying a net charge, andconsequently receiving a small shock on openin~ thedoor.

Figure 1: Induction Charging

(:<?= '-<.-

---.

.='

(a) (b)

Material Assessments

In tribo-charging the precise nature of the two surfacesis very important in determining the level of chargeacquired. For a material to be moderately charged it oftenmeans that there is only one electron too many or toofew for every billion (109) molecules, or more. This meansthat very low levels of impurity, or seemingly insignificantvariations in materials, may have a very dramatic effecton thei(chargj\lg properties. As a result, when assessingpotential problems this disproportionate effect leads toa need to look at all but the simplest of materials on acase by case basis, often by carrying out measurementsunder carefully controlled laboratory conditions to assessthe propensity for charging in a particular situation. .

I~

No material is a perfect insulator so that even ascharge is being generated it will also be dissipated byconduction. The actual charge which will be observedwill therefore be the result of the dynamic equilibriumbetween charge generation and dissipation rates. Therate of charge dissipation depends on the conductivityof the material and, just as various factors affect chargegeneration, this too can vary- Especially important is therelative humidity of the air around the material. Minutequantities of water adsorbed on to a surface from thesurrounding air can have an enormous effect on

StClticEleGfricily;Glflcf ~ ~\Ij .' "".,' "",.. ,,'

the ~RhGlrt11QGl!ut,i~~.I)I:nd\~s.~"'.~", '11 -","

conductivity, dissipation rate and hence detected chargelevels.

Static Discharges

Hazards (as opposed to problems) due to static almostinvariably arise because sparks, or discharges, are ableto ignite flammable materials. In fact, several types ofelectrostatic discharge have been identified, each withcharacteristic energies.

For example, a brush discharge will occur from aninsulating surface. The energy in a brush discharge isrelatively low; often below the limit of perception if it isto a person in a typical working enviwnment.Nevertheless, a brush discharge will ignite manycommon solvent .vapours.

----

A capacitive discharge (or spark) dissipates energy,E, according to the relationship:

E = ~CV 2

2

V\1hereC is the capacitance of the conductor on whichc~arge was stored, and V the voltage to which it wasraised. Capacitance is dependent upon geometry andIqcation, but simplistically increases with the size of the

~

ject. Assuming a potential of 10kV (easily attainable)ble-2 shows capacitances and spark energies available

f om some common objects.III

~bjectI

Table 2: Capacitances and Spark Energies forSome Common Objects

TypicalCapacitance

(pF)10 -20

10 -10050 - 300200 - 300100 - 1000

10kV SparkEnergy

(mJ)0.5 - 1

. Q.5- 52.5 - 1510 - 155 - 50

mall metal items (eg scoop)

mall containers (eg bucket)ledium containers (eg drum)uman body

~arge plant (eg reaction vessel)I .

I Other types of discharge are corona, cone and

!

r

.

opagatingbrUSh discharges. Each has a characteristicnergy range associated with it. Corona occurs fromharp points and is a low energy type of discharge, evenompared with brush discharges. Cone discharges occurcross the conical surface of powder, and especially

wanules and pellets, as they collect in hoppers, silos~nd other containers. These are more energetic thanbrush discharges and can ignite common solvent vapoursand some dusts. Propagating brush discharges are the

most energetic of all. They arise when chargeacc.umulates on two sides of a thin insulating layer. Apractical scenario where a propagating brush dischargecan occur is on an insulating plastic liner used inside afibreboard drum.

Systematic Hazard Assessment

We have seen the ways in which charge may begenerated. Whether or not it actually happens in aparticular situation will be a matter for prediction from

. laboratory measurements or direct measurements onplant.

If charging occurs we have seen how it could lead todifferent types of discharge. The energy which might beavailable from such discharges can be predictedaccording to its type and, in some cases, specificmeasurements on site.

In general, the risk is ignition of a flammable material:solvent vapour, permanent gas or dust. Whether or notthis could occur will simply depend on a knowledge ofits sensitivity to ignition (or Minimum Ignition Energy)and comparing that with the potential discharge energy.For many simple materials ignition energies arepublished in the literature. For special materialsdetermination of Minimum Ignition Energy is a routinetest for laboratories such as Chilworth Technology's.Photograph 1 shows a small dust explosion initiated bya spark of known energy as part of the series of testscarried out to determine Minimum Ignition Energy.

Photo 1: Minimum Ignition Energy test.

Chemical News,January 2009 I 37I!

-"'-

Avoiding Hazardous Situations

With the risk, and the mechanism for its arising,properly understood there are a number of options foravoiding the hazards, although in practice one optionoften stands out as the most appropriate.

It may be possible to avoid charging in the first placeby altering the process or operating conditions in someway. Alternatively, charging may have to be acceptedand charge accumulation prevented. There is no excusefor allowing charge to accumulate on conductors; theycan always be earthed. However, as we saw earlier,although earthing the plant is crucial and will preyentcharge accumulation there, it will generally have no effectat all on the material inside which is being processed.

In the end, it is sometimes the case that if the desiredprocess is to be carried out charging cannot be avoided,the risk of discharge cannot be avoided and a flammablematerial is at risk of ignition from the discharge. If this isthe case the only option will be to avoid ignition byinetting.

Incidents from the Pharmaceutical Industry

It was stated at the beginning of this article that theph~rmaceutical industry was pa,iicularly vulnerable toele$trostatic hazards and problems. In the light of theforElgoing discussion it is now possible to explain andjust,fy that statement.

flant is often deliberately and necessarily operatedun er low humidity conditions. For many materials thisme ns they are at their least conductive and thereforemost susceptible to charging.

Materials are mostly organic, and often chemicallyvert active. Experience in Chilworth Technology's testlab6ratories is that, increasingly, pharmaceutical productsare amongst the most sensitive to ignition.

In common with most other processes ever higherspe~ds are required to maximise the plant capacity andmin,mise cost. As discussed earlier, higher transferspeeds, and more energetic processes in general, leadto higher levels of charge.

The cleanliness and law cost of pl<;lsticsas containerstherpselves, liners for other containers, liners for plant,or plant materials means that there seems to be nostopping their increasing use. These, of course are the

38 11 Chemical News, January 2009

"""y

""~

;ik

"'.-

very materials which will lead to many of the problemsdiscussed earlier.

Often floors (and walls) are finished with materialseminently suitable for washing down and maintainingcleanliness but with little thought to their condUctivity.People and objects moving around on an insulating floorhave no means of dissipating charge and will becomecharged.

Many of these problems arise in other industrysectors. However, few others bring them all together inthe way the pharmaceutical industry does. Theconsequences can be graphically illustrated with a fewreal examples from Chilworth Technology's archives ofincident investigations. So~e of the hazards are soobvious with hindsight that it is difficult to see how they

.could have been missed - yet they were. Others requireconsiderable insight even when the concepts are fullyexplained.

Dust Explosion - Sieving

Operators were scooping powder from an FIBC tothe sieve unit. The fines product from the sieve weredropped into a stainless steel bin. One day there was anexplosion and fire in the bin accompanied by thick blacksmoke. Fortunately the operators were unhurt and ableto evacuate the room before returning to extinguish thefire. Following a detailed analysis of the incident andmeasurement of relevant variables an explanation forthe incident was found.

'\ The bin was on insulating wheels. Although an"earthing lead was provided it was not used. A simulationexperiment demonstrated that the powder would becharged by the sieving operation, and carry charge intothe bin. The powder was found to have a MinimumIgnition Energy of 15mJ. Given the measuredcapacitance of the bin, and the rate of chargeaccumulation in the bin based on the experimentallydetermined powder charge, it would have taken a littleover half a minute to reach the point where a dischargewould be sufficiently energetic to ignite the powder. Allthe ingredients were there but why did the incidentactually occur this time. It turned out that as the bin filledthe centre of gravity moved until it was able to rockforward about its larger central wheels. When placed injust the right (wrong!) position this caused it to contactthe main body of the sieve. To avoid dust in the room asa whole the bin had been covered with a plastic sheet.This caused a concentrated dust cloud to escape where

--""" ~ --

the sheetwasdrapedoverthe sieveoutlet- exactlywherethe spark occurred on this one occasion when the binrocked and madecontact with a goodearthafterit hadacquired sufficient charge to provide an incendivedischarge.

Seemingly insignificant changes had been made toworking practices, until all the ingredients forthe incidentwere in place. It then simply needed the bad luck ofcoincidentally bringing them all together at once. Andthe solutions were simple enough. Improve the binearthing, preferably by the use of conducting wheels,Provide local dust extraction so that concentrated dustclouds are avoided.

Reactor Charging 'Explosionr--

,--..

A 45001 vessel had. been washed with acetone andleft to dry overnight. The next day drums of a powderintermediate were manually tipped into the vessel viaan open manway. After drurn number six was added there

wfls an explosion and a fireball enveloped the operators.

T~is incident did cause serious injury and resulted inlost production, compensation payments and, less

t~ngibly? loss of workforceconfidence.I

~

The investigation found the powder in this case to havea ignition energy of between 1.and 5 mJ. Trials showedt at the morning after an acetone wash the concentration0 acetone vapour could be about 50% of the lowere plosible limit. Clearly the acetone on its own would noti nite, but it could contribute to a hybrid flammable mixturet gether with the flammable powder. It was found that the0 erators footwear and gloves were both insulating. Thisi entifiedtwo possible isolated conductors - the drums andt e operators. Simulation experiments showed 1OkVcould

e attained by the drum during emptying. Given its

~

.apacitance this meant that a 1OmJspark could have beenroducedfrom the drum (or indeed from the operator).learlythiswas the sourceof ignition.I

iIn this case a number of solutions are possible. The

essel could be inerted and double valves or flap valvessed. Local dust extraction may also have improved the

~ituation. However, probably the most important

E

commendations, whether or not the others werei plemented, was to provide dissipative footwear and

loves and an earth clip for the drum.IIgnition of Hexane Vapour

A hexane-Iadenpowderwas beingtransferredbytwo

-- - - --j---

~

..

~

operators from a centrifuge buggy lined with a woven.cloth "buggy bag" into plastic lined fibreboard drums.When most of the powder had been removed one of theoperators grabbed the "buggy bag" on one side and pulledit from the supporting hooks. The ignition occurred as aflash fire on the "buggy bag" which immediately spreadto the nearby fibreboard drums.

All conducting plant, and the operators were wellearthed in this case. However, the buggy bag and thedrum liners were insulating. The powder was found' tobe very insulating, and hexane known to be. Althoughnot discussed earlier, insulating liquids can generatecharge in a special case of tribo-charging - known as astreaming current. The insulating process materials,"buggy bag" and liners provided the means for chargegeneration. In this case hexane has an ignition energy

.of only 0.24mJ. That means a bwsh discharge from aninsulating surface, in this case the "buggy bag", was themost likely cause of the ignition.

The recommendations here were to change the finalwash solvent to, preferably a non-flammable liquid but,at least a conducting one. In addition all insulating plasticswere banned from the area. .

Electrostatic Discharges over Toluene

An operator observed blue flashes in a stirred vesselcontaining toluene as the solvent. Although undernitrogen this observation was, to say the least,disconcerting.

A special probe was constructed and installed on sitein order to carry out measurements during a trial run inwhich operational parameters could be varied. It wasfound that below a certain temperature the level of chargerose dramaticaily, falling away again as the temperaturewas raised. This temperatl,lre was coincident with thetemperature at which solute started to come out ofsolution.

It is well known that a multi-phase liquid system.charges very much more highly than a single 'phase.The problem was solved by ensuring the temperaturenever dropped below the experimentally-found transitiontemperature. .

Vacuum Dryer Incidents

A vacuum dryer had suffered several productdecompositions at the end of the drying cycle leading to

Chemical News, January 2009 39

Statlc . E.>.,'v~; ',' ...,.

tt,,8":9ha,~j''. '"

over-pressurisation of the vessel and, in some cases,loss of containment.

Chilworth Technology undertook a wide-rangingresearch project and discovered that the product wassusceptible to acquisition of high charge levels. However,the key was in showing that an electrostatic phenomenonknown as the Paschen effect could occur in chargedpowders - a fact hitherto unrej)orted. This meant thatrelieving the vacuum could cause electrostaticdischarges that, inturn, could initiate the decomposition.

The solution was to limit the level of vacuum used.This ensured that breaking the vacuum at the end of thecycle no longer caused discharges as a result of thecharge gained during drying.

Conclusions

The pharmaceutical industry is particularlysusceptibletal electrostatic problems in general, and hazards in

Qnd't~' ".'~<

particular. However, a proper understanding of therelevant phenomena and appropriate physical propertyinformation allows a systematic approach to definingoperational practices and plant designs to minimise therisk.

Incidents, such as those described earlier, can beeffectively and systematically investigated. However, thedirect and indirect costs of allowing any avoidableincident to occur can be very significant and it is clearlymuch better to undertake an expert assessment of aplant and its operations prior to any incident. Then, byadopting appropriate recomrnendation$, the risks canbe minimised or avoided altogether.

. Best of all would be to consider the full implicationsof potential static hazards (alongside other hazardsassessments) before the plant is even off the drawing ~

board. Increasingly companies recognise thisthemselves, but it is probably true to say that many havehad to learn the hard way. ...