Weighing Performance during Typhoons Improving Balance Stability

ScinoPharm Ltd.

Environmental disturbances cause issues with balance stability around the world. Peak typhoon season in particular creates significant weighing challenges for users of sensitive balances - measurement issues that standard balance filters simply cannot compensate for. Rather than accept this adverse effect on weighing results, METTLER TOLEDO set out to find a solution that could be utilized during Application Note such disturbances. The result is reliable weighing performance in extreme weather conditions.

A typhoon is a mature that develops in the northwestern part of the Pacific Ocean between 180° and 100°E, commonly referred to as the Northwest Pacific Basin. This region includes all the territory north of the equator and west of the International Date Line, including the South Sea, and it is the most active basin on Earth, accounting for almost one-third of the world's annual tropical cyclones.

While no official “typhoon season” exists, storm activity in the Northwest Pacific Basin tends to be highest from August to October, a period which nearly corresponds to Atlantic hurricane season. Along with high storm frequency, the basin also features the most intense tropical storms on record. Features like warm sea temperatures, atmospheric instability, high humidity, air rotation, and low ver- tical wind shear combine to create these potentially deadly storms.

Aside from the obvious danger of typhoon landfall, manufacturers in Pacific regions such as coastal China, , , Taiwan, Vietnam, South , and The —and, to a lesser extent, Southeast Asian nations such as Malaysia, Thailand and Singapore—have reported poor balance stability and difficult weighing operation during periods of typhoon activity. 2

METTLER Note Application Application Note TOLEDO Table 1:Tropical ScalefromRegionalSpecializedMeteorological Center(RSMC)Tokyo CycloneIntensity meant that all potential improvements to filter settings couldbetested for robustnessbeforeimplementation. improvementstofiltersettings allpotential meant that field.This approach beforedeploymenttothe solutions, ofpotential was usedforofflineanalysisandtesting balanceweighingcellandenvironmentsensorsbyanonboardindustrialPC.Thisrawdata fromthe directly pressure.Upto1.8gigabytes(GB)ofraw datacouldbecollectedperday anddifferential perature, vibration, atmosphericpressure, tem- humidity, task. Themeasurementboxalsocontainedsensorswhich measured the ideallysuitedforthe areas, madethen frequent usageofWXS205balancesintyphoon the features, alongwith inindustrialdevices.These installation forthe acompactdesign destined automated weighing,combinedwith downto10microgramsduring"on-board" offeranimpeccablerepeatability sion balanceswerechosenasthey contained fourWSX205SDU/15DualRangebalances(Fig.1).Thesehigh-preci- special measurementboxthat To ateamofMETTLERTOLEDOweighingexperts createda weighingandrobust datacollection, ensuresensitive Establishing of Instability Cause the world. tumultuous coastalregionsaroundthe tobalanceusersin didit,andwhyitmatters Here’showthey undertropicalstormconditions. hing operations ofweig- mostsensitive requiredforeventhe repeatability createsthe ultimately time, ger balancestabilization lon- inaslightly despiteresulting studywasdevelopmentofaspecific“Typhoon that, filter” The endresultofthe developingandmaturestorms. events,includingpressurechangescausedbyboth typhoon joint studyinvolvedmeasurementofaseriesatmosphericparametersduringrecorded sothe ce instabilities, balan- buildingscausingthe ofthe wasbasedonwindandatmosphericpressure-inducedvibrations hypothesis ingredients(API).Theinitial pharmaceutical research anddevelopmentcompanymanufacturerofactive ScinoPharmTaiwan disturbances,METTLERTOLEDOcollaboratedcloselywith Ltd.,aleadingprocess weather phenomenonanddevelopmeasurestoimprovebalanceperformanceduringstrong this To investigate further forPartnering Advancement in Research steep finesfromregulators. canincur offinalproducts—andproductirregularity efficacyandsafety where evenminuteerrorscanaffectthe balancesinregulatedindustriessuchaspharmaceuticals cularly seriousforoperatorsrelyingonhigh-resolution - canbeparti Theseweighingdifficulties canmeandaysorevenweeksofbalanceinstability. peak season,this orgreater(Table PacificBasinacquiretropicalstormstrength 1).Duringanormal Northwest cal cyclonesinthe 20-30tropi Annually,typically stormscanbecostly. Periods ofabnormalbalanceperformancecausedbythe Tropical Depression Tropical Storm Storm Tropical Severe Typhoon Category ≤ 61≤ km/h knots 33 ≤ km/h 88 - 62 47knots - 34 -11789 km/h knots 63 - 48 118≥ km/h knots 64 ≥ winds Sustained 3

METTLER Note Application Application Note TOLEDO Figure 2:Typhoon EventsinPacificduringJune-Dec 2013 Figure 1:Measurementboxhardwarelayout 2013). September (16-24 TyphoonUsagi and 2013) July (7-14 TyphoonSoulik during collected data on only focus will article affectedTaiwan redboxesinFig.2,this effects (Fig.2).Whilealleventsthat areindicatedbythe typhoon study to opportunities offered events tropical of number a season, 2013 the During E. 120°16’55.7” buildinginShan-Hua,Tainanupper floorofatwo-story andlongitudeof23°07’11.5”N –atalatitude labatScinoPharmTaiwan, Quality locatedonthe The measurementboxapparatuswasplacedinthe 4

METTLER Note Application Application Note TOLEDO Figure 3: Comparison of Fluctuations duringTyphoonsFigure 3:ComparisonofAtmosphericPressureFluctuations SoulikandUsagi be concluded absolutely that this degradation in performance was solely pressure or vibration based. inperformancewassolelypressureorvibration degradation this be concludedabsolutelythat itcouldnot although impactoftyphoons, poorbalanceperformancetothe ofthe showed astrongcorrelation collecteddata events.Analysisofthe typhoon datesofthe againstthe chartplotted finalstandarddeviation the pressure duringeacheventdecreasedbyanorderof-20mbar. alsoincreased,asshownby Pressurevariability eventsshowedseveraltrends.AsevidencedbyFig.3,atmospheric The atmosphericdatacollectedduringthese pond to the passing by of the typhoon. typhoon. passing by ofthe pond tothe corres- ofupto5mgwererecorded,whichdirectly weighingsignal.Deviations onthe typhoon impact ofthe the show 5) and 4 (Figs. events both during signal weight filtered the of fluctuation the displaying Graphs for10days. od wasalsoplotted same 15minuteperi- weighingsignaloverthe ofthe balance.Thestandarddeviation ofthe weighing capacity hing signal(inmilligrams)over aperiodof15minutesatzeroload,halfandfull load inrespecttothe maximumminusminimumdifference inweig- everyhourfor10days.Thisvalueisdefinedasthe was plotted balanceweighingsignalwasseverely disrupted.Thebiggestweighingsignalchange the typhoon, During the Typhoon Usagi 5

METTLER Note Application Application Note TOLEDO Figure 5: Disruption ofWeighing SignalsduringTyphoonFigure 5:Disruption Usagi TyphoonSoulik during Signals Weighing of Disruption 4: Figure Typhoon Soulik Typhoon Usagi 6

METTLER Note Application Application Note TOLEDO Figure 6: Initial Typhoon Filter Approach TyphoonFilter Initial 6: Figure activity. by on typhoon brought vibrations and changes atmospheric of face in the filtering accurate repeatable, provide and balances stabilize would the that a solution find to Toledo needed Mettler step? next the was So, region. what Pacific in the users balance for problem real a presented results the facilitate, they weighing of type the and balances high-resolution of nature the on Based The green point shows the release of a stable value for measurement when using the Typhoon filter. the using when measurement value for astable of release the shows point green The Typhoon conditions. value under astable reach will never balance astandard that observe clearly you can Here 2013. in September Usagi (red line) "Typhoon (blue improved filter" line), the versus Typhoon settings of filter passing the balance during standard the between ismade Acomparison balance. the onto placed are weights the when level the to med iszoo data the clarity For weights. internal balance the with cycles load minute one three time. Fig 7shows settling balance and results weighing of repeatability between compromise acceptable an as established fully care were parameters filter optimum The tested. were stages filtering additional productivity), (which decreases time settling increased against repeatability improved the evaluating and filter initial the of results the on Based 2: filtering Testing stages. additional Refining Approach threshold. acertain below changes load to sensitive less was it bances, distur to relation in averaging original the than robust more proved approach new this Fig.However,6. whilst in indicated as signal weighing the of smoothing by successive dictated stages involved multiple structure filter initial taken. The was outset the at filter astronger applying of approach values, the filtered averaging of Instead filter.” 2: Develop “typhoon aunique Approach investigated. was approach alternative an reasons, these For implementation. for a terminal for need the and commands, dedicated of use the to due impractical proved solution However, the case, in this ship. alarge board on small samples weighing or vibrations; frequency low emits which equipment facturing manu near weighing are situation this of Examples issue. an cause can conditions in environmental ations vari large where applications, dedicated other in several applied successfully been already had method This time. of over afixed period collected 1: data filtered average to DevelopApproach algorithm an Taiwan ScinoPharm for in 2014.tion taken: were steps following The solu afinal, to workable led optimizations and approaches of aseries operations, balance abnormal reported historically that environment in an events typhoon after and during, before, collected data continuous Using Identifying Optimal the Solution ------7

METTLER Note Application Application Note TOLEDO Figure 8: Repeatability ofWeighing MeasurementduringTyphoonFigure 8:Repeatability SoulikusingvariousFilterMethods StandardbalanceFilter, new"Typhoon"Figure 7:Comparisonofthe filter versusthe The repeatability of the Typhoon filter was also significantly improved. significantly also was Typhoon filter the of repeatability The 8

METTLER Note Application Application Note TOLEDO to pay for significantly enhanced stability, repeatability, and accuracy. and stability, repeatability, enhanced significantly for pay to asmall price be to proving times settling longer slightly results—with weighing of quality high same the ensure to on switched be can filter typhoon the events, However, weather severe during measurements. precise and fast usual the with manufacturer pharmaceutical the providing continue can operation normal activity, typhoon significant without year the of times other at that means required when mode typhoon the enable to ability The software. robust environmentally more the of advantage take to Typhoon filter the Taiwan enable only ScinoPharm events, need Now, mode. typhoon check-weighing during the into incorporated filter typhoon 2014July improved in the service with into put balance XPE205test single is, a except that test—none, calibration internal an pass to Taiwan enough stable ScinoPharm at were balances analytical no disturbance, the During Sea. China South 2014 the in Kalmaegi September Typhoon in during field the in test the to put finally was filter stronger the 2014 season, the to typhoon start quiet relatively a After Field-Testing “Typhoon the Solution” 2014. September since manufactured balances analytical XPE the in standard is software This software. in METTLER TOLEDO’s in place now are filters XEAnalyticalBridgeV1.10 stronger results, lab improved nificantly sig and users balance from feedback positive on However, open. based isstill technically two the of bination acom or pressure effects, by vibrational induced are disturbances whether of question the so completed, been yet not has but initiated was data seismic recorded actual with experiment the from obtained data of Correlation tions. vibra typhoon-like random by the impacted negatively not was performance balance in place, "Typhoon filter" the with that determined was it laboratory, in the obtained data the as well as themselves, events real-world the from taken data ashaker. on on aWXS205 using module Based METTLER TOLEDO laboratory in the test simulated also were conditions vibrational typhoon-like themselves, events the during obtained data with Along affected. isdetrimentally performance balance the where adegree to layers, sediment on especially vibrations, the of strength the increase locally to seems typhoon a of However, presence weak. the isvery nearby. landfall effect make this Usually themselves typhoons not or whether of regardless instruments, weighing sensitive of performance the impair to potential the with vibrations earth random into translate can and world in the omnipresent are vibrations ocean Micro-seismic Vibrations Good Table ResultsUsingDifferentFilterMethods 2:Repeatability New tuned Typhoon filter 20 seconds over Filter, averaged Standard 10 over seconds Filter, averaged Standard Filter Method Used (see Typhoon 25of new filter Table the using 2): by afactor improved was filter, repeatability the standard balance the with Compared filtering. data of methods different three of repeatability the Soulik. Fig. 8compares typhoon during recorded was repeatability The Green Red Yellow 8 Fig. on Plot x 25 x 15 x 10 x Improvement Repeatability - - - Application Note Global Switzerland MarCom © 06/2015 AG30267103 Mettler-Toledo Subject totechnical changes +41Fax 3060 944 44 Tel. +41 22 11 944 44 Schweiz Greifensee, CH-8606 &Laboratory Weighing Technologies AG Mettler-Toledo www.mt.com For moreinformation