5History of Photometers & HEPA Filters “You can’t have one without the other”
6The photometer was first since there was a need David Sinclair, Ph.D.(Nephelometer)(Not portable and had logarithmic display)A Nephelometer is a type of photometer identified by Dr. David Sinclair. The need was monitoring of air pollution, particularly in cities where steel was manufactured.
7Air Pollution Research 1933 Cities such as Pittsburgh and Baltimore. This is a picture of the ship building located in Baltimore, MD during the 1930’s and the steel foundries in the background.
8WW II generated a high priority need It all started in 1942
9Some Key PlayersWendell Anderson, Humphrey Gilbert, Dr. Melvin First
10Captured WW II German Gas Masks The US had performed no gas mask development since WW 1German masks used cellulose asbestos media patented by Dräger Werke in 1933Wendell Anderson working with H&V developed media comparable to the German sampleGerman masks were acquired by the US and found to have far superior filtration characteristics to those used anywhere else in the world. Given the chemical warfare agents at the time, this became a very high priority for the US Military.
11M10A1 CanisterThe M10A1 respirator canister was the first high efficiency filter deployed by the US Military.
13Smoke Penetrometer to test the gas mask paper & filters Now that a high efficiency paper (media) and filter existed, there was a need to test them. The Q-127 was the product developed to do this. There are still Q-127 instruments utilized today in both military and commercial industries. ATI was the premier provider of this instrument since the early 1960’s and still services components of this instrument today. This instrument is no longer manufactured and has been replaced by the TDA-100P instrument.
15Space FilterUS military needed filtration of room areas for people and used gas mask media for a pleated filter for larger air flow ( )Know as “Collective Protection”Used cardboard spacers between pleats, which had high air flow resistance
17Manhattan ProjectHumphrey Gilbert safety engineer at Los Alamos was sent to Oak RidgeFilters used in Manhattan Project were very thick with extremely high pressure dropForesaw need for high efficiency air filters in HVAC systems
18Absolute Air FilterGilbert (now with AEC) unhappy with Army Space filter and its limitationsAEC in 1948 gives Author D. Little a contract to redesign filter and find a supplierWalter Smith Ph.D. comes up with corrugated cardboard separator idea
21HEPA Filter Development First Air Cleaning Seminar for AEC personnel held June 1951 at Harvard Air Cleaning LaboratoryFirst Handbook on Air Cleaning distributed at meetingFires at several weapons plants made need for high temperature and water resistance necessary
22Second Air Cleaning Conference held in Ames, IA Melvin W. First Ph.D. Presented four research papers
23Military develops procedures and Apparatus to test High Efficiency filters and related productsPenetrometersQ-127 Low FlowQ-76 Med. FlowQ-107 High FlowRough HandlingWater Repellencyetc.
24HEPA Development for Fire & Water Resistance (1953)
31Next Portable Photometer TDA-2A (1964) Cabinet covered in white Formica(White Rooms)First Scanning ProbeThe portable photometer migrated to testing in place filters in what was then known as “White Rooms”, the precursor to the clean room as we know it today. This also created a need for scanning the filter to find and repair defects, which resulted in the first scanning probe.
32Next Portable Photometer TDA-2B (1965-66) Ergonomic front panel design
33First Commercial Standards (1967) Up until now, the commercial world knew nothing of high efficiency (HEPA) filters, how they were tested for efficiency or tested in place. This migrated into the Pharmaceutical and life sciences, aerospace, and semi-con/micro-electronics industries we are familiar with today.
37Before Photometry… What is a Filter? How to Test a Filter Filtration MechanicsMPPSHow to Test a FilterEfficiencyLeak (Integrity)
38What is a Filter? Separation of one phase from another Solids or liquids in Air (Home Furnace)Solids in liquid (Auto Oil/Gas)Liquid in Air (Air oil separators)Liquid in liquid (RO)Gas in Gas (Activated Carbon)
39What is a Filter?A Controlled Leak! Filtration controls the amount of impurity that is allowed to pass. There is no such thing as a ‘perfect’ filter (no resistance and 100% efficiency)
40Filtration Types Materials Performance Fibers Fibrous Structures MicroporesPerformanceHigh EfficiencyLow EfficiencyFilters can be classified by many other means: applications, other properties such as temperature limitations, etc.
41What are we dealing with? Solids or liquids in Air&Fibrous FiltersOther structures are more common in ultra filtration and special applications
42Fibrous Filters Typically Depth Filters Fibrous Filter Construction Filtration throughout media depthVery low to very high particle removalFibrous Filter ConstructionFilter Disks/PadsPleated CartridgesBags/PocketsMembranes are usually considered surface filter as opposed to depth filters. Due to low loading requirements, this is less relevant in high purity applications found in Clean manufacturing.Filter cartridges can be cylindrical or flat panelsIn practice, multi stage filter or media used to get desired loading/performance effect.
43Filtration Mechanics Pressure Drop Single Filter Efficiency Collection Mechanisms
44Darcy’s LawAs systems require maximum resistance ratings, filter manufacturers must design around this parameterΔP
45Darcy’s LawThe greater the flow rate through the filter the greater the resistance. Not rocket science, but what is that relationship?
46Filter Pressure Drop CLEANROOM HVAC Pressure Drop Manufacturers can design filters for specific applications i.e. known flow rate and pressure drop requirements. But what about the filtration characteristics?Linear Pressure Drop Inertial Velocity(Darcy Law) Regime Regime
47Filtration Assumptions Fibers are in cross flowParticles are ‘collected’ upon contact with fiberUniform fiber diameterUniform particle size
48Single Fiber Efficiency Air Volume swept out by fiberSingle Fiber Efficiency ηs =Particles collected by fiber Particles in volume of air geometrically swept out by fiber
49Fiber Structure - HEPAImage of a typical fibrous media. There are many fibers and many layers of fibers. There is variation of fiber to fiber thickness and thickness within fibers. The basic filter model uses the particle capture by one fiber and does not consider interference from additional fibers or additional fiber layers.
51Particle Collection: Sieving Filtration: Theory, Design & Practice I21 Feb 2002Particle Collection: SievingFibersParticle FlowSieving is the most intuitive collection method. However, this mechanism usually happens by pre-filters not HEPA filters.ATI Technical School
52Particle Collection: Interception Filtration: Theory, Design & Practice I21 Feb 2002Particle Collection: InterceptionInterception is simply a collision. The particle impinges upon the fiber and is collected. Interception is greater for larger particles simply because of the geometries involved and reduced diffusion effects.ATI Technical School
53Particle Collection: Inertial Impaction Filtration: Theory, Design & Practice I21 Feb 2002Particle Collection: Inertial ImpactionLarger particles, due to their greater mass and, thus, greater inertia, separate from the air stream, as the air stream bends around a fiber, and impacts upon the fiber. Inertial effects are greater for larger particles and at higher flow rates.ATI Technical School
54Particle Collection: Diffusion Filtration: Theory, Design & Practice I21 Feb 2002Particle Collection: DiffusionDiffusion occurs due to the intrinsic natural random movement of small particles. Diffusion is much more significant for small particles, their random motion brings them in contact with the fiber and are collected. Diffusion effect reduces with larger sizes.ATI Technical School
55Particle Collection Mechanisms Van der Wall’s Law is what holds a particle to the fiber once contact is made. These are the small electrical bonds at a molecular level and is how a gecko is able to climb vertical and upside down without suction or adhesion properties.Electrostatic Deposition can also happen when a fiber is charged and the particle is charged opposite that of the fiber (i.e. opposites attract). This is only relevant for some media. To achieve any permanent benefits from electrostatic effects, the media must remain dry and charged. If moisture is present the charge leaks away. Realistically, the high efficiencies attained by electrostatic media are typically quite transient and dissipate quickly.
56High Efficiency Filter MPPS (Most Penetrating Particle Size) Filtration: Theory, Design & Practice I21 Feb 2002High Efficiency Filter MPPS (Most Penetrating Particle Size)ImpactionDiffusionMPPS 0.08 to 0.18µmPenetrationThe addition of all collection mechanisms leads to a Penetration Curve with an unique peak or max penetration. As the particle size increases, the diffusion effects decrease and the interception and impaction effects increase leading to the peak penetration. By adding the red and blue curves shown above, we obtain the yellow curve whose peak identifies the particle size at the maximum penetration. This is called the Most Penetrating Particle Size (MPPS).Keep in mind that the efficiency η = 1 – penetration. These two terms are often used interchangeably to define filter performance.It is intuitively obvious that efficiency at particles larger than the MPPS will be higher (or lower penetration). It is less obvious that the efficiency is also greater at particles smaller than the MPPS!! In other words, the MPPS is the worst case for a filter. More on the factors affecting MPPS in the next lecture.Particle SizeATI Technical School
57Filtration is Selective Particle Collection 0000Small and big particles are more effectively collected
58Impact of Flow Rate on Efficiency As you decrease the flow rate from the specified flow rate, the % penetration will decrease (i.e. efficiency will increase); however, the MPPS will shift upward. The converse is also true. Typically, as a filter loads and the resistance increases, the flow rate will gradually decrease, shifting the MPPS upward, but the filter efficiency improves!
59Filtration: Theory, Design & Practice I 21 Feb 2002Media Velocity is Lower Than Face VelocityMedia VelocityFace VelocityA Filter’s Performance is Determinedat Media VelocityMedia Velocity = Face Velocity x Face Area / Media Area in FilterMore media area the lower the pressure dropATI Technical School
60Consideration in Filter Design Penetration or efficiencyCritical Particle SizePressure Drop or resistance(Cost)
61Filter Tests Efficiency Leak/Integrity Media Manufacturer Filter Manufacturer3rd Party CertificationLeak/IntegrityField/In-Situ
62Filtration: Theory, Design & Practice I 21 Feb 2002Efficiency TestingChallenge filter with aerosol at or near the MPPSThis requires a mono-dispersed aerosol (tight distribution)Measure upstream and downstream concentrationSequential or parallel (simultaneous) measurementsEfficiency = 1 – DownstreamUpstreamNote: This is a ‘global’ measurement (i.e. entire filter)Efficiency testing is a simple concept: challenge a filter and measure what gets through. It is a ratio-metric measurement (i.e. Percentage) based on known parameters: flow rate and upstream challenge. Although simple in concept, there are always complexities which we will discuss later.ATI Technical School
63Typical Filter Efficiency Test System Filtration: Theory, Design & Practice I21 Feb 2002Typical Filter Efficiency Test SystemNote the original MilStd 282 testing methodology (i.e. photometry) has been omitted from this diagram, although it is still a viable methodology for efficiency testing per IEST RP It is important to note, the use of photometry in this manner is the same methodology specified and employed by most media manufacturers and specifically by both the Nuclear and High Efficiency Respirator industries.Recommended filter performance test system from IEST RP-CC-0021ATI Technical School
64Filtration: Theory, Design & Practice I 21 Feb 2002What is a Leak?Leak is a local measurement.Leak is generally 5 –10 times the average penetration at a local spot.Therefore we do not require a mono-dispersed challenge aerosol at the MPPSEasier and less costly to generate larger particles with a broader distribution.Keep in mind that there are variations in the performance of media due to the random nature of the manufacturing process of fibrous media.ATI Technical School
65A Good Filter Will Allow Few Particles to Penetrate Filtration: Theory, Design & Practice I21 Feb 2002A Good Filter Will Allow Few Particles to Penetrateo o o o o o o oo o o o o oo o o o o oo o oo o o o o o o o00o o o o o 0oFilter allows a small fraction of particles to pass throughATI Technical School
66Filtration: Theory, Design & Practice I 21 Feb 2002Add a Leak and…o o o o o o o oo o o o o oo o o o o oo o oo o o o o o o o00o o o 00 o o o o o 0 0O o o o o o o 0 0o o oNow, if there is a hole (leak) in the filter, many particles go throughATI Technical School
67Test for a Leak/Integrity Filtration: Theory, Design & Practice I21 Feb 2002Test for a Leak/IntegrityMeasure local penetration at each sample location to determine:Filters are not faulty / not been damagedFilters have been installed properlyThere are no leaks in the mounting frame / between mounting frame and housingSystem contains no by-pass of the filterLeak testing is a local measurement; however, by scanning the entire filter and where it is ‘sealed’ into the overall system, the entire filtration system integrity is tested and validated.ATI Technical School
68Leak/Integrity Scanning Filtration: Theory, Design & Practice I21 Feb 2002Leak/Integrity ScanningScanning is the method where a probe sweeps the entire surface of a filter and its sealing mechanism looking for local penetration. Typically the probe samples at the same velocity as the air flow in the filter.There are various instruments that may be used as the detector and we will discuss these later. The speed of the scan rate also affects the ability to detect a leak and is addressed more fully in IEST CC RPEnsure that all parts of filter are within specificationATI Technical School
69Efficiency vs. Leak/Integrity in General Terms Efficiency TestingLeak TestingGlobalLocalDetermine % PenetrationYesYes1Pinpoint LeaksNoManufacturer TestYes2Field TestWe will review this in greater detail later, but for now, it is important to note that efficiency testing is to qualify the filter itself and Leak testing is to ensure the integrity of the filter, its housing, and the filtration system as a whole.Also, it is important to note that some manufacturers can/will perform a ‘Leak’ test of the media (i.e. scan test), but typically charge extra for this. This guarantees any variability in the media will not exceed the leak threshold in the field which may not be picked up with an overall, ‘Global’ efficiency test.1 – Local % penetration, not overall efficiency2 – Due to variations in the media manufacturing process
72Let’s Review… Particulate penetration as Size of interest? However….. EfficiencyLeakSize of interest?MPPS from 0.08 to 0.18 micronEven when looking for a leak?However…..
73What is a particle?Particles come in all shapes and sizes. Also, visible organisms can change shape a grow complicating things even further.So what is it we refer to when specifying a particle size?
74The Simplest Particle?The real world is very complicated, so we try to simplify it as best we can. The simplest 3-dimensional shape to deal with is a sphere. And a sphere’s defining measurement is its radius or diameter (2r).
75Equivalent Particle Size Diameter of a Sphere that has the same magnitude of a chosen Property as the particle In questionOptical Diameter (Eye/microscope)Scattering Diameter (Light Scattering)Electric Mobility (Charge On Particle)Mobility (Diffusion)Stokes Diameter (Drag Forces)Aerodynamic Diameter (Settling Speed)When we say “Particle Size” it is useless unless we also state how we measure it, it is not a unique measurement.Particle size as commonly noted implies spherical particles and their optical diameter, but this can be and usually is a big misconception.Also, if the particle size refers to its physical size, then a denser particle will settle faster than a less dense particle of the same size. Think of dropping a pound of cotton vs. a pound of lead. In commercial terms, we typically load a truck by volume, but pay by weight.So we need to be specific when referencing particle size. In this business we generally speak of scattering diameters and mass.
76In Filtration, Collections of Particles Interest Us Aerosol Solid/Liquid and Gas (dust, fog)Hydrosol Solid/Liquid and Liquid (milk, paint, lotion)Foam Gas in Solid or Liquid (sponge, shampoo)Aerosol is a collection of many particles. If they are all the same size, then it is termed mono-disperse. If they are multiple sizes, then it is poly-disperse.
77Key Concepts Drag Force – proportional dp > 10µm Slip Correction for dp < 10µmStopping Distance and Relaxation TimeSettling velocityMobility and Electric MobilityDiffusion and Brownian MotionCoagulationDimensionless NumbersThese are key properties we deal with when discussing and measuring particles and aerosols.Dimensionless numbers? They are very common in all engineering and sciences. They relate two complimentary or competing forces and eliminate the need for units.For example, when the viscous drag on a particle balances the flow of the fluid it is traveling in, this equates to the Reynolds number, Re. Re is a fundamental concept in all fluid dynamics. The term ‘Laminar Flow’ which we all have heard is a condition that occurs at low Reynolds numbers.
78Units of Measurement in Filtration µm Micrometer, 10-6 meterNm Nanometer, 10-9 meterA Angstrom Unit, meter1 µm = 1,000 nm = 10,000 ASome perspective:1cm is 1000 µm 1in = 2.54 cm = 25,400 µm
79Approximate Particle Diameter Particle Sizes of InterestItemApproximate Particle DiameterEye of a Needle1,230 micronsBeach Sand100 – 2000 micronsTable salt~100 micronsHuman hairmicronsTalcum powder~10 micronsTobacco Smoke0.01 – 1.0 micronsBacteriamicronsVirus<0.005 – 0.05 micronsMany of these particles can be a chain or a collection of smaller particles.Cigarette smoke can be a chain or a ring.For inhalation, small particles behave like a gas and go in and out of the lung freely and, although are inhaled, do not settle in the lung to be absorbed. The larger ones get caught in the nose or fall past the nose to the ground. It is those particles in between (approximately um) that are of interest from a respirable perspective.It just so happened that DOP aerosol as generated to test filters back in the 50’s and 60’s had a 0.3um (Mass Mean Diameter)…a lucky break of nature…
80Typical Particle Sizes 0.3umSo why are we concerned with viruses? They require a transport mechanism…mucus, saliva, phlegm…all aerosolizes when someone coughs or sneezes.
82Typical Settling Velocity Diameter (um) Feet/minThe particles we are interested in tend to stay airborne for quite some time. Larger particles will settle very quickly.
83How is an Aerosol Characterized? Consider this data setFrequencyX (range)Raw DataNormalized0 to 5100205 to 6706 to 86030Here, the count is the number between the X range. There are 100 counts between 0 and 5.Since the X ranges are not equal, we normalize the data by dividing the counts by the X range. Therefore the normalized count in the 0 to 5 range is equal to 100/(5-0) or 20.The result is shown graphically in the next slide.
85Size Distribution Terms Independent variable, XMean XmStandard Deviation, σMode (peak)Median (50th percentile)For our business, the X variable is almost always the Particle diameter…the equivalent diameter!
86Arithmetic Standard Deviation This is the fundamental distribution taught in high school, commonly called the ‘Normal’ distribution. It is also known as the Gaussian or Standard Distribution. Some refer to it as the ‘Bell’ curve.Although simple to use, it requires the underlying data to meet very strict conditions of independence and normality. Believe it or not, most natural occurrences and even those man-made, do not meet these criteria.
87Geometric Standard Deviation The world is geometric. The symbol ‘ln’ is the natural logarithm.If X = ey, then Logarithm e X = Y or the common logs.In simple terms, if one substitutes X = ln x (or log x), the resulting data follows the Standard distribution.
88Particle Size Distribution Most Natural Processes are Geometrically Distributed and often multi modalNumber (count), Surface, or Volume (mass) Distribution are common weightingsThe data can be weighted in any manner. Number, surface area, or volume (mass) are common.
89Why is this important?Due to data weighting… …the 0.3um DOP particle measured in the 50’s is a volume (mass) mean… …whose number (count) median is actually closer to 0.18um…the MPPS
90Geometric vs. Arithmetic Consider Geometric as Arithmetic or Linear on a log scaleGeometric spreads out the lower endIn Arithmetic, equal differences result in equal spacingIn Geometric, equal ratios result in equal spacing
91Geometric vs. Arithmetic Geometric scales are very useful when the data spread is very large. It helps enlarge the scale at the lower end.If one wants to divide the range into different intervals, we use equal differences in arithmetic scale and equal ratios in a geometric scale.Equal intervals Arithmetic: = = … (equal differences)Geometric: 10/7.48 = 7.48/5.62 = … (equal ratios)
92Why Use Log Normal Distributions? No Negative ValuesFor known distributions, the Means of different weightings can be calculated from the othersFor example, Mass (volume) Mean calculated from Count (number) mean
93Number and Volume Weighting dpNumber CountsVolume (πd3/6)11,000,000523,00010100
94Key Points to Remember Particles come in many shapes and sizes ALL measurements determine some physical property and provide an Equivalent SizeIn filtration, larger particles settle out and are not important, particularly cleanroomsGeometric Distributions (aka Log Normal) are used to define particle size distributionsMass and Number weightings are common in particle measurements
98LIGHT PARTICLE INTERACTION RamanFluorescenceReflectionDiffractionScatteringAbsorptionIncident LightRefractionAll materials absorb lightAbsorption is the negative of reflection and, in technical terms, it is also referred to as the imaginary part of the refractive index.A black body is good absorber. No light comes out and hence it appears blackEmission
99Common PropertiesRefraction – the apparent change in direction of light due to change in refractive index within one medium or between dissimilar mediumReflection – redirection of light at the surface of a materialDiffraction – bending or deflection of light around a particlePolarization – oscillations of the light occurring in a defined planeRefraction is dependent upon the refractive index of the particle/medium.Polarization implies that the light waves are oriented in one plane.
100Complex Numbers (4)1/2 = 2 What is (-1) 1/2 ? It is defined by the symbol iA complex number is written as (a - b i) where the imaginary part is represented by the symbol i
101Refractive Index Refractive Index of a material is a complex number Usually given as 2 - 4iThe imaginary part is due to absorption
102Refractive Index of Common Materials Quartz iGlass 1.5 to iPSL iCa Sulfate iCarbon 2.0 – 0.33iIron – 1.63iThe refractive index is a complex number. The imaginary part is denoted by i and denotes the absorptive component of the refractive index.Most clear materials such as glass, oil drops, etc., have no imaginary or absorptive component of the refractive index. Many metals and colored materials absorb light.
103Refraction Refractive Index = sine r / sine i i r Refractive index is a ration of the velocity of light in a vacuum to that in the medium or particle. The imaginary part is proportional to the absorption index and the wavelength of light of interest. The refractive index is a complex number denoted as R – iA, where the imagininary part is denoted as iA.Refractive Index = sine r / sine i
104Size and Scattering Regimes Size Range of InterestA Hydrometeor is any water or ice particles that have formed in the atmosphere or at the Earth’s surface as a result of condensation or sublimation. Water or ice particles blown from the ground into the atmosphere are also classed as hydrometeors. Some well-known hydrometeors are clouds, fog, rain, snow, hail, dew, etc.
105Elastic ScatteringRedirection of Incident Light without change in wavelengthRefraction – internal to particle, wavelength and composition dependentReflection – at surface of particle, dependent on wavelength and compositionDiffraction – external to particle, independent of wavelength and composition
106Particle Size Affects Elastic Scattering Optical Particle Size (α) α = πd / λ where:d = particle diameterλ = wavelengthScattering Intensity (Is)Is = λ2 f(α) where:f(α) is a size dependent functionSimply put Scattering Intensity is a function of both the particle size and the wavelength of light it is subjected to.
107Light ScatteringParticles much smaller (< 0.025um) than the wavelength of light results in Rayleigh ScatteringParticles comparable to the wavelength of light (0.025> x < 2.5um) results in MIE ScatteringMuch larger particles result in geometric scattering.Simply put Scattering Intensity is a function of both the particle size and the wavelength of light it is subjected to.
108Light ScatteringParticles much smaller (< 0.025um) than the wavelength of light results in Rayleigh ScatteringParticles comparable to the wavelength of light (0.025> x < 2.5um) results in MIE ScatteringMuch larger particles result in geometric scattering.Visible Light is between 300 to about 600 nm or (0.3 to 0.6 um). The scattering intensity varies as the inverse of the fourth power of the wavelength of light.Most of what every photometer and particle counter manufacturer does is in the MIE scattering realm.
109Rayleigh ScatteringParticle diameter is much, much less than the wavelength of light.Scattering equal in forward and backward direction.Intensity increases as fourth power of 1/wavelength of lightThis is why the sky is blue – blue light has a much smaller wavelength than red.
111Mie Scattering Wavelength, λ Incident light Particle size approaches wavelengthForward scattering more intenseThe intensity of scattering varies less than the fourth power of 1/wavelengthThis is the regime we operate in particle instrumentsIncident light
112MIE Scattering Particle size approaches wavelength Forward scattering is more intenseThe intensity of scattering varies less than the fourth power of 1/wavelengthThis is the regime we operate in with regard to particle instruments.
115Multiple and Single Particle Sensing and Sensors
116Multiple Particle Sensing Sampling Volume ≥ 1/number concentrationIndependent of sample volume (measuring a ‘cloud’)Precision depends on averaging time (due to variations in the ‘cloud’)Remember, 1/number concentration has the same units as volume.
117Multiple Particle Sensors Extinction (umbrella)Smoke Meter (soot content in exhaust stack)Transmissometer (visual ranging)ScatteringIntensity (nephelometry)Backscatter (LIDAR)PhotometryLIDAR is light detecting and ranging. It is used for remote sensing of aerosol by back scattering.
118Photometer Aerosol is illuminated by a light source Total scattered light is detected by PMTTotal concentration is measured
119Filtration: Theory, Design & Practice I 21 Feb 2002Photometer0 00 0 00 0AerosolFocus OpticsCollector OpticsPhotodetectorElectronicsLight SourceThe photometer is calibrated with a known aerosol size, distribution, and concentration, DOP or PAO. It uses internal reference settings to accommodate other aerosol substrates.Instrument response is instantaneous and constant.Output response is based on mass level of aerosol – this is gravimetrically validated as part of the calibration process.Output magnitude dependent upon mass concentration of aerosol in sampling chamber.ATI Technical School
121NephelometerParticle density is a function of the light reflected into the detector from the illuminated particlesScattered light is a measure of the total scattering in the view volume.Larger volumes than in aerosol photometers.Used for ambient measurements (pollution and climate) and visibility measurements.
122Multiple Particle Sensors Summary Light-Particle interaction results in scatteringOptical instruments in particle measurement is dependent upon the particle size and scattering propertiesMultiple particle scattering is independent of volume; depends on averaging time.
123Single Particle Sensing Sample Volume << 1/number concentrationRequires a precise, known volume of sampled airSame as counting events; precision depends on total countsRemember, 1/number concentration has the same units as volume
125Filtration: Theory, Design & Practice I 21 Feb 2002Particle CounterLight SourceFocus OpticsCollector OpticsPhotodetectorOOElectronicsA Particle Counter assumes all particles have the same physical characteristics as the material used to calibrate the instrument (i.e. PSLs)Light scattered by each particle is counted and its magnitude measured. The output is based on the size (or effective cross sectional area) of a discrete particle. The quantity of pulses is the particle count and magnitude the particle size.The size measurement is calibrated with various mono-dispersed poly styrene latex spheres.ATI Technical School
126Common Particle Counter Challenges Multi Valued ResponseCoincidenceProblems at small particle sizes
127Multi Valued SignalThe scattering intensity response is not unique for small particle sizes. That is, particles of two sizes can give the same response making particle detection by size difficult as smaller sizes.
128Coincidence O Photodetector Electronics Light Source Focus Optics Collector OpticsPhotodetectorOOElectronicsConsider two small particles entering the light beam. They both scatter light at the same time. This coincident scattering results in a scatter equal to a larger particle. Hence coincidence will count small particles as a larger particle but at t lower concentration.Coincidence is common at higher concentrations. The cure is to dilute the sample.
129Small Particle Detection Scattering intensity is very smallCommon SolutionsIncrease intensity of lightChange light source wavelengthNew techniques?
130Small Particle Detection Increase IntensityImproves signal from particlesIncreases noise from air moleculesReduce the wavelengthShifts the curves to smaller sizesIncreases noise from moleculesReduce the viewing volumeThere are many ways to reduce noise, both electronic and physical. Reducing the viewing volume is the easiest; however, the sample volume is critical in single particle sensing!
131Condensation Nucleus Counter Filtration: Theory, Design & Practice I21 Feb 2002Condensation Nucleus CounterCondensation NucleiiSaturator VaporParticles are introduced into an environment with saturated vapor, typically alcohol, although there are models using H20 and other liquids. The vapor condenses on the aerosol much like cloud seeding in artificial rain These droplets grow in the condenser to larger sizes and are detected by traditional light scattering.Since artificial clouds or drops are created around particles, particles as small as 0.03 µm can be detected. However, the formation of drops around particles makes it impossible to determine the original size.AerosolAlcoholATI Technical School
132Single Particle Sensors Summary Most optical measurements are in the MIE regimeSingle particle counting requires known volume of sampled airSame as counting events; precision depends on total counts – long sample times at low countsProblems of coincidence at high concentrationsNon unique response and low signal to noise ratio at small sizesSmall sizes handled by using smaller wavelengths or proprietary methods
133Photometer vs. Particle Counter vs. CNC Measures Total AerosolResponse Linear With Total Aerosol VolumeRequires Known Aerosol And Relatively High ConcentrationProblems:No Particle Size InformationRequires High Concentrations
134Photometer vs. Particle Counter vs. CNC Particle Counter (Laser)Detects and sizes particlesCounts by sizeMeasures down to 0.1 μmCan use any aerosolProblems:Assumes everything measured is a PSLMulti Valued ResponseCoincidenceProblems at small particle sizesLong sample times to obtain a statistically valid results
135Photometer vs. Particle Counter vs. CNC Particle detector onlyCan measure less than 0.05 μmProblems:Requires mono-dispersed aerosolCounts all particles, noise at bottom endLong sample times to obtain a statistically valid results
138What is a Standard?A KNOWN and Universally Accepted value of a physical property or quantityMeter for measure of length◦C for Temperature
139Why are Standards Necessary? They establish accuracy of measuring instrumentsCalibrate the accuracy of instruments in useTemperatureScalesFlow MetersPhotometersParticle Counters
140How Does this Affect Aerosols? A Standard Aerosol has KNOWN propertiesParticle Size and DistributionConcentrationShape (usually spherical)Chemistry (inert)And Refractive Index in our businessQuite often it is an instruments response to Concentration which we are calibrating. Mass for photometers and counts for particle counters.Care must be taken when diluting particles as particles can behave differently than gases and losses can skew readings and data. For sub micron particles, we are fortunate that this is not much of a problem unless there are charging issues at play or non typical factors at play. There are always nuances that must be considered.
141Aerosol Standards Mono-disperse Aerosol Near Mono-disperse Aerosol ‘Single’ size particles σg ≤ 1.4Instrument Calibration for PCsResearch & DevelopmentNear Mono-disperse AerosolNarrow distribution 1.4 < σg ≤ 1.6Some Production QC Level of accuracyPoly-disperse AerosolBroad distribution 1.6 < σgIndustrial measurementsInstrument verificationSome instrument calibrationsATI uses both near mono-disperse and poly disperse aerosols. Since Photometers respond to the total mass of the aerosol, poly-disperse is not a big problem and, in fact, provides a more stable result.
142Mono vs Poly DisperseMono disperse has a sharp peak and narrow spread
143Mono vs Poly DispersePoly-disperse aerosol has a large σg typically > 2. Flatter peak and much broader spread.
144Implication on Size Dispersion For a log normal distribution95% of the particles are betweenMean σg2σg2 & Mean
145Implication on Size Dispersion An aerosol with mean of 0.3umσg = 1.295% of particles are between 0.21 & 0.43umσg = 2.095% of particles are between and 1.2um
147Three Types of Aerosol Mono-disperse: σg ≤ 1.4 Near Mono-disperse: 1.4 < σg ≤ 1.6Poly-disperse: 1.6 < σgThere is no hard fast definition that is accepted in industry; however, from an application perspective and reviewing various global filter testing standards, this schedule tends to hold true.
152Laskin Nozzle Concentration Calculations Output of Laskin Nozzle is defined:By # 20 PSIG Concentration = # Nozzles x 13,500Total Flow (cfm)By # Jets at 20 PSIGConcentration = # Nozzles x 3,375*Note: there are 4 jets in a standard Laskin NozzleA Laskin Nozzle was selected by the Naval Research Laboratory as its output is mostly sub-micrometer and the aerosol size/distribution and concentration was very repeatable and reproducable.
153Laskin Nozzle Concentration Calculations You can calculate the system concentration, if you know:System Air Volume (CFM)Number of Laskin Nozzles/Jets at 20 PSIGYou can calculate the number of nozzles/Jets, if you know:Desired system concentration
154Wright NebulizerThe common aerosol can is similar to this type of nebulizer, except the compressed air/gas is inside the can and is released when the spray button is depressed. This is also the type of aerosol generator used for inhalers.
155Pneumatic Nebulizer Compressed Air Aerosol Liquid The high speed jet bubbling through the liquid creates the bubbles in the liquid that generate the particles at the surface. Very similar to water splashing in a boiling liquid. The small particles or droplets are entrained in the air and carried out. The large droplets fall back into the liquid.Liquid
156Thermal Condensation Aerosol Generators O 0 O o 0Quench AirVaporAerosolLiquidCompressed GasHeaterThe quench air condenses the vapor into drops. The quench air rate and temperature can have an impact on particle size/distribution.Quench Air
157Thermal Aerosol Generator PolydispersedProduces a greater level of aerosol concentration than pneumatic type nozzleApplications include higher flow systemsMedian particle size is smaller than pneumatic generationOutput concentration cannot be calculated as output is variableSize and distribution shift with concentration
158Ultrasonic NebulizerSimilar to the Pneumatic (Wright) Nebulizer except the agitation to produce the drops is created by a vibrating crystal at an ultrasonic frequency.
159Spinning Disk Particle Generator Size and concentration of the aerosol can be controlled by the liquid feed rate and the rpm of the spinning disk.
160Exploding WireThe exploding wire generator works much like a spark plug in a car, except that the electrode disintegrates by the discharge of the capacitor. The vapors condense to form the aerosol. Rather exciting, but not very practical and not commonly used.
161Test Dusts Standard Dust (Arizona Road Dust) Mainly Silica with Mass Mean Diameter of 7 µmσg of 3.6Specific Gravity of 2.7Originally collected from Arizona Desert
162Test Dusts ASHRAE Custom blend of: 72% ISO , A2 Fine Test Dust,23% powdered carbon5% milled cotton lintersAttempt to simulate natural dust for HVAC
163Test Dusts SAE Dusts – automotive filter testing Fine Coarse Mass Median Diameter ~ 25µmNo particles > 100µmCoarseMass Median Diameter ~ 60µm10%+ may be larger than 100µm
165PSL Aerosols NIST Traceable PSL Particles Examples Methodology Atomize PSL in Liquid (water)Evaporate the liquidNIST Traceable PSL AerosolPSL Aerosol is the ONLY NIST traceable standard particle size
166PSL AerosolsPSL is always atomized from a suspension in water or other liquid. The water is removed in the heater/dryer leaving PSL particles.
167Residue & PSL Aerosols Impurities in water become small particles These particles can be counted as particles in standard aerosol, especially by a CNCIt is always a good practice to use clean water that has been filtered and distilled as the carrier.
168Residue & PSL AerosolsAssuming a typical 2um atomizer droplet and 10 ppm purity water, Residue particle size can be computed 10 x 10-6 = (residue dia/drop dia)3 Residue Diameter = 0.04umResidue particles are especially a problem when generating small PSL sizes. The residue is not much of a problem when using particle counters since they are not sensitive to such small particles. It becomes a big problem when using CNCs.
169Residue ParticlesThere can be more residue particles than those of interest; however, they are extremely small and their respective mass will be very small as well.
170PSL Aerosols Common for Cleanroom Applications NIST traceable PSLs are expensivePSLs are easy to aerosolize, but output concentration is variableLimited by residue at small sizesNot available in large sizesPSL gives excellent optical responseUsed as calibration aerosol for particle counters
171Electric Mobility Classification Electric Mobility Classification exploits the physics of how movement of charged particles is affected in an electric field. The particle moves right with the fluid flow and down due to the electric field.Only a small fraction of particles with the predetermined electric mobility will arrive at the opening and get collected.For a given flow rate and electric field, the mobility range is determined by the width of the slit and the width of the aerosol flow.As shown, the upper mobility (or smaller particle size) is defined by the particle at the out edge of the air flow which enters at the upstream edge of the opening/slit. Likewise, the upper end of the particle size is defined by the particle at the inner edge of the flow which just enters the opening at the downstream edge.The aerosol extracted from the opening has a narrow and predetermined mobility range and will be mono-disperse.
172Points to RememberStandards are required to verify and calibrate instruments and devicesA Standard aerosol can be either poly or mono disperseR&D standards are more precise and are for laboratory useIndustrial standards are easier to produce and are widely used
173Points to Remember cont’d. Poly-disperse aerosols are commonly generated by atomization, nebulization or mechanical meansOnly a few techniques are available to generate very tight, mono-disperse aerosolsNIST traceable PSLs are generated in small quantities and are very expensivePSL concentration output is variableResidue particles can be a problem in small sizes
176What is a “Filter” Testing Standard? A DOCUMENTED and Universally Accepted method of obtaining a Qualitative Performance Measurement
177Why are Standards Necessary? They establish consistent methodologyProvide a guide for understanding and compensating for variables encountered during practical application
178How Does this Affect Filter Testing? A Test Standard defines methods and limitsAerosol characteristicsSystem Operating ConditionsTesting protocolsAllowable Challenge ConcentrationsSampling RateMaximum “Allowable” LeakageScanning Speed
179Standards & Recommended Practices Organizations AACC (American Association for Contamination Control)ISO (International Organization for Standardization)BNL (Brookhaven National Laboratory)ASTM (American Society for Testing and Materials)EN (European Norm)IEST (Institute of Environmental Science & Technology)ANSI (American National Standards Institute)ASME (American Society of Mechanical Engineers)DOE (Department of Energy)
180Filter Testing Standards CS-IT (1968) Standard for HEPA filtersCS-2T (1968) Standard for Laminar Flow Clean Air Devices - Installation Leak Test (filter and gaskets)CS-2T (1968) Standard for Laminar Flow Clean Air Devices - Induction Leak Test (seams and joints)(2005) Cleanrooms & Associated Environments, Annex B- Test Methods (Informative)IH62300 (2001) In-Place HEPA Filter testing, Section 6.2 EquipmentEN High Efficiency Air Filters (HEPA & ULPA)-Part 2: Aerosol Production, Measuring Equipment, Particle Counting StatisticsIES-RP-CC001 HEPA & ULPA FiltersIES-RP-CC006 Testing CleanroomsIES-RP-CC007 Testing ULPA FiltersIES-RP-CC034 HEPA & ULPA Filter Leak TestsFed Std 209E (1992 by IEST) Most referenced in “Filter” industry was replaced by ISO & 2 in November 2001NSF 49:2008 (Annex A:2008) Biosafety Cabinetry: Design, Construction, Performance & Field Certification-Performance TestsNSF 49:2008 (Annex F:2008) Biosafety Cabinetry: Design, Construction, Performance & Field Certification-Field TestsASME N509 (2008) Nuclear Power Plant Air-Cleaning Units & Components)N510 (2007) Testing of Nuclear Air Treatment SystemsN511 (2007) In-Service Testing of Nuclear Air Treatment Heating, Ventilating and Air-Conditioning SystemsDOE-HDBK-1169, DOE Handbook Nuclear Air Cleaning Handbook, Section In-Place Testing
181Industries Using Photometry PharmaceuticalNon-pharmaceuticalCivilianNuclear PowerMilitaryNuclear WeaponsChemical WeaponsBiological Weapons
182Photometer vs. Discrete Particle Counter 1968 CS-1T Standard for HEPA FiltersPhotometerUpstream challenge aerosol must be at least 27ug/lMaximum Leakage = 0.01%Scanning rate = 2 inches per 1 inch from filter faceParticle CounterNo defined method
183Photometer vs. Discrete Particle Counter 1968 CS-2T Standard for Laminar Flow Clean Air DevicesInstallation Leak Test (for filter & gaskets)PhotometerUpstream challenge aerosol must be at least 27ug/lMaximum Leakage = 0.01%Scanning rate = 2 inches per 1 inch from filter faceParticle CounterNo defined method
184Photometer vs. Discrete Particle Counter 1968 CS-2T Standard for Laminar Flow Clean Air DevicesInduction Leak Test (for seams & joints)PhotometerAmbient aerosol must be at least 10 E3 above FFMaximum Leakage = >FFScanning rate = 2 inches per 1 inch from joint or seam within clean zoneParticle CounterAmbient aerosol must be >300K particles/ft3Maximum leakage >100 countsScanning rate = 2 inches per 1 inch from joint or seam within clean zone
185Photometer vs. Discrete Particle Counter 2005 ISO Cleanrooms & associated controlled environments, Test methodsInstalled Filter System LeakagePhotometerUpstream challenge aerosol of between 20 & 80ug/lMaximum Leakage = 0.01%Scanning rate = 2 inches per 1 inch from filter faceLimitationsEfficiency < MPPSOil aerosols allowedAbility to achieve required concentrationsParticle CounterUpstream challenge aerosol (Too much detail to list 6 pages) sufficiently high that Np >2 & <10Maximum leakage = 0.01%Scanning rate ≤ 3.14 inches per second (varies depending on probe dimensions, Np, sample rate
186Photometer vs. Discrete Particle Counter 2005 ISO Cleanrooms & associated controlled environments, Test methodsContainment TestPhotometerUpstream challenge aerosol of between 20 & 80 ug/lMaximum Leakage = 0.01%Scanning rate ≤ 2 inches per second at 2 inches from joint, seal or mating surfacesLimitationsEfficiency < MPPSOil aerosols allowedAbility to achieve required concentrationsParticle CounterThe greater of:ambient count X E103> 3.5 E106 particles/m3Maximum leakage = ≤ ambient count X E10-2Scanning rate ≤ 2 inches per second at 2 inches from joint, seal or mating surfaces
187Photometer vs. Discrete Particle Counter 2001 IH62300:2001 In-Place HEPA Filter testingPhotometerUpstream challenge aerosol 104 greater than ambientMaximum Leakage = 0.03%Duct measurement-No scanParticle CounterNo defined method
188Photometer vs. Discrete Particle Counter IEST-RP-CC006 Testing CleanroomsPhotometerUpstream challenge aerosol of between 10 to 20 ug/lMaximum Leakage = 0.01%Scanning rate = 2 inches per 1 inch from filter faceParticle Counter3 x E108 particle size of interest (10 counts per Appendix B, Exp. 1)Maximum Leakage = 0.01%Scanning rate =[(Cc)(Ls)(Fs)(Dp)]÷ [(60)(Np)]Result in or inches per second (Appendix B, Exp. 1)
189Photometer vs. Discrete Particle Counter IEST-RP-CC034 HEPA & ULPA Filter Leak TestsPhotometerUpstream challenge aerosol of between 10 to 90 ug/lMaximum Leakage = 0.01%Scanning rate = 2 inches per 1 inch from filter faceParticle Counter2.8 x E108 particle size of interest (10 counts per Appendix G, Exp. 1)Maximum Leakage = 0.01%Scanning rate = [(2.8)(108)÷1000][(0.0001) (28.3)(1.25)÷(60)(10)]Result = in or 0.5 in/sec(Appendix G, Exp. 1)
190Photometer vs. Discrete Particle Counter 2008 NSF 49 Biosafety CabinetryPhotometerUpstream challenge aerosol of at least 10 ug/lMaximum Leakage = 0.01%Scanning rate = 2 inches per 1 inch from filter faceParticle CounterNo defined method
191Points to Remember!Photometer & particle counter results are not likely to correlate due to the different weighting of the technology used.Photometry response is mass weighted while particle response is number weighted.Filter testing process is the same regardless of photometer or particle counter use.Measure upstream challengeMeasure downstream penetration by scanning or point samplingCalculate penetration
192Points to Remember!Filter testing process complexity varies between photometers and particle counters.Challenge concentrationPhotometers: Variable range of ug/l with ug/l being typical (Std defined)Particle counters: Variable range of 3.0 E105 to 3.0 E108 (Std defined, calculated to achieve desired Np)Measure DownstreamPhotometers : 0.01% leakage maximum while scanning rate of 2 inches /second in most cases (Std defined)Particle counter: 0.01% leakage maximum in most cases while scanning at a calculated, statistically, determined rate (Std defined)Particle counter challenge requires diluter to prevent coincidence error/optics saturation
193Points to Remember! Calculate Penetration Photometer: Ratio of upstream challenge to downstream expressed as a percentParticle counter: Ratio of upstream challenge to downstream expressed as a percent
194Points to Remember!Each technology has design strengths and weaknesses which decide where its use is “reasonable” in filter leakage testing.PhotometerAerosol generator for upstream challenge “oil” aerosolNo diluter necessary for Upstream measurementsConsistent & essentially “Calculation Free” test methodLimited to systems with efficiencies ≤99.997%Typically consistent results among multiple unitsRobust “core” technologyParticle counterAerosol generator required in most cases, but at a lower outputUse of solid or liquid aerosol possibleAbility to test systems to % efficiencyDiluter required for upstream samplingUnit to unit result consistency difficult to achieveMore sensitive detection system results in less-robust instrumentDiluter adds additional errors to measurement