Compressor Instability with Integral Methods Episode 2 Part 3 ppt

mix-7.2 Emission of Air Sound There are four major sources of air sound generated during blast cleaning operations: r sound emitted from the pressure generating unit compressor, engine a

Health, Safety and Environment

7.1 Safety Features of Blast Cleaning

7.1.1 General Safety Aspects

General aspects of health, safety and environment (HSE) for blast cleaning cations are summarised in Fig 7.1 ISO 12944-4 states the following for surface

appli-preparation in general: “All relevant health and safety regulation shall be observed.”

Blast cleaning owns an injury potential General sources of danger to blast cleaningoperators include the following:

r reactive forces generated by the exiting air-abrasive mixture (see Sect 6.6.4);

r hose movements;

r uncontrolled escape of pressurised air;

r damaged parts being under pressure;

r dust and aerosol formation;

r sound emitted from equipment and blasting jet;

r impact from rebounding abrasive material and debris from the impact point

It is generally recommended to carry out a risk assessment of the actual ronment where a blast cleaning job will be done before starting the job This riskassessment may include (French, 1998):

envi-r how access is to be gained?

r is there a need for scaffolding?

r is there confined space?

r what is the surface like where the operators will have to stand?

r the availability of day light or artificial light;

r the presence of electrical supplies/equipment;

r nature of contaminate: Is it toxic? Is it a pathogen? Is it asbestos based? Is itharmful or corrosive?

r general layout that will allow visual contact between the blast cleaning team;

r permit requirements;

C

 Springer 2008

uncontrolled TPM emissions (kg/yr)

uncontrolled TPM

emission factors

(kg/kg)

dispersion modeling

toxicity ratings for

efficiency

of APCD (%)

controlled PM10emissions (kg/yr)

total inhalation-induced cancer = Σ UREi* Conci &

non-cancer risk = Σ Conci/RfCi

controlled metal emissions:

Cr, Mn, Ni, Pb (kg/yr)

metal fractions in TPM

Cr, Mn, Ni, Pb (kg/kg)

Fig 7.1 HSE risk analysis for blast cleaning processes (Kura, 2005)

r safety of access (e.g working on motorways or hazardous areas such as ery where flameproof equipment and earthing to avoid static electricity may berequired);

refin-r who or what will be affected by flying debris?

r is noise a problem?

r will containment be necessary?

r where will the effluent go? (for wet blast cleaning and slurry blast cleaning)

In that context, ISO 12944-4 states the following: “Personnel carrying out face preparation work shall have suitable equipment and sufficient technical knowl- edge of the processes involved.”

sur-7.1.2 Risk of Explosion

Some source of explosion during blast cleaning can be electric discharge sparks.Safety hazard analyses identified that static electric charges occur in the followingthree circumstances:

r small particles flowing through piping;

r small particles passing through fine filters or nozzles;

r abrasive particles impinging fixed parts

Table 7.1 Results of spark measurements during blast cleaning (Stuvex Belgium)

Dry blast cleaning Wet blast cleaning Voltage at the blast cleaned surface in V 500 2–3

Voltage of static electricity at the nozzle in V 5,000–10,000 0

Results of spark measurements performed on oil containers with dry blast andwet blast cleaning techniques are listed in Table 7.1 Dry blast cleaning generatedhigh levels of voltage at the blast cleaned surface as well as at the nozzle The use

of wet blasting equipment helped to keep these levels low Elbing (2002) reportedabout measurements of the electrostatic charging of steel during the blast cleaningwith carbon dioxide pellets The author measured values as high as 3,000 V, and

he found that the discharge current increased with an increase in air pressure andstand-off distance The discharge current was rather high for shallow impact angles,but reached a lower saturation level for impact anglesϕ > 50◦.

The electrostatic discharge in hose lines due to friction between hose walland flowing abrasive particles can be managed through the use of blast cleaninghoses with low electric resistance Values for the electrical resistance lower than of

103⍀/m are considered to allow a safe charge elimination (BGR 132, 2003)

The effects of impinging abrasive particles on the ignition of explosive gas tures are not well understood Dittmar (1962) highlighted the fact that sparks mustcreate a certain temperature field in order to ignite gas mixtures Duration and in-tensity of the temperature field determine the danger of explosion The author citedexperimental results from machining operations, and he reported that the forma-tion and subsequent combustion of small metal chips generated temperatures up to2,300◦C These high temperatures were much more critical than the temperaturesreached at the tool–chip interface during the material removal process (see Fig 5.17for the situation during blast cleaning) Smaller chips generated higher temperaturesthan larger chips Dittmar (1962) also reported that rusted steel substrates were muchmore sensitive to spark generation compared with clean steel substrates

mix-7.2 Emission of Air Sound

There are four major sources of air sound generated during blast cleaning operations:

r sound emitted from the pressure generating unit (compressor, engine and powertransmission);

r sound emitted from the abrasive air jet travelling through the air;

r sound emitted from the erosion site;

r sound emitted from accompanying trades

Two other items of concern are noise generated by air supply in the helmet andsound attenuation of the helmet

State-of-the-art air compressors are regularly equipped with sound insolatinghoods or even placed in containers Thus, the air sound emission is limited up to70–75 dB(A) More critical is the air sound emitted by the jet This noise is gener-ated due to friction between the high-speed jet and the surrounding air as well asdue to turbulences Thus, the sound level depends on the relative velocity betweenjet and air, and on the surface exposed to friction Consequently, air sound levelincreases as compressor pressure, nozzle diameter and stand-off distance increase.This is verified in Fig 7.2 where the effect of the nozzle pressure on the noise level

is shown The noise level increased almost linearly with increasing air pressure.Equal trends have been reported for the noise emitted during dry blast cleaning withcarbon dioxide pellets (Elbing, 2002)

Fig 7.3 illustrates results of measurements performed at different blasting sites,where dry blast cleaning, shot blast cleaning and wet blast cleaning were applied.Figure 7.3a includes results from measurements at a dry blast cleaning site Theactual blast cleaning application generated the highest noise levels Figure 7.3b and

d shows results from measurements at wet blast cleaning sites The noise ated during the actual wet blast cleaning application was lower than the noise levelmeasured for the dry blast cleaning in Fig 7.3a Figure 7.3c contains results frommeasurements at a shot blast cleaning site The noise level was again lower thanthe noise level for the dry blast cleaning process mentioned in Fig 7.3a, which wasdue to the facts that no air was involved in the mechanically driven shot blast clean-ing process, and that the blast cleaning head was sealed It can be recognised that

gener-Fig 7.2 Effect of compressor pressure on noise level (Schaffner, 1997)

1 - dry blast cleaning

2 - scaffolding

3 - maintenance air supply system

4 - picking up solid waste (grit,paint)

5 - transportation of solid waste

Fig 7.3 Results from noise-level measurements during steel surface preparation jobs (Knipfer and

Funke, 1997) (a) Dry blast cleaning; (b) Wet blast cleaning; (c) Shot blast cleaning; (d) Wet blast

cleaning

Fig 7.4 Noise levels at different locations (Ognedal and Harbak, 1998)

the actual blasting operations (dry blast cleaning, shot blast cleaning and wet blastcleaning) generated the highest noise levels among all trades Shot blast cleaning(which works with shrouded blasting tools) and wet blast cleaning are comparativelysilent Noise emission can notably be reduced if shrouded or sealed tools are used.Figure 7.4 shows results of air noise measurements performed inside the helmet

of blasters Ognedal and Harbak (1998) concluded from these measurements thatblast cleaning may create loss of hearing to the workers, if no additional hearingprotection is provided

The permissible air noise level depends on the exposure time This is illustrated

in Fig 7.5 based on regularity limits stated in the German standard ‘BGV B3 L¨arm’

It can be concluded from the graph that ear protection equipment must be worn byany personally involved blasting cleaning operator (see Sect 7.8)

7.3 Emission of Body Sound

Body sound characterises waves, which carry noise and travel through solid als Therefore, even if windows, doors, etc are properly closed to lock out airbornenoise, persons may any way experience certain noise levels This noise is generateddue to vibrations; they occur during the tool impact and depend on the acousticproperties, especially on the sound velocity and the acoustic impedance, of both

materi-Fig 7.5 Critical exposure times for different preparation tools (solid line according to BGV B3

L¨arm; points from different sources)

the material to be subjected and the preparation tool The evaluation parameters

of the vibration are its amplitude and velocity (frequency) No measurements areavailable from dry blast cleaning operations However, there are some measure-ments available from concrete facades treated with other surface preparation tools.Amplitudes and vibration velocities generated by the tools are plotted in Fig 7.6.The two mechanical methods generated rather high values for vibration velocities,whereas the application of water jets led to low vibration velocities

7.4 Emission of Dust

A mist of paint debris and broken abrasive particles is generated during dry blastcleaning in the immediate environment of the operator An example is shown inFig 7.7 A simple model for the evaluation of dust during the blast cleaning ofmould casings was introduced by Engelberg (1967)

Unfortunately, the dust is difficult to control The only way to prevent it is theuse of shrouded tools Another way to protect the operator is the application ofmechanically guided tools or robotic machinery Anyway, both methods fail as itcomes to the cleaning of complex structures A reduction in dust exposure is pos-sible by adding water to the air particle flow (wet blast cleaning and slurry blast

Fig 7.6 Measurements of body sound emitted from different surface treatment tools (Werner and

Kauw, 1991);veff – effective vibration velocity; s – vibration amplitude Preparation methods:

1 – water jetting, 2 – hammer and chisel, 3 – jack hammering, 4 – pneumatic hammer, 5 – angle grinder

Fig 7.7 Dust formation during dry blast cleaning (Photograph: Muehlhan AG, Hamburg)

Fig 7.8 Additional working time in a shipyard due to dust formation (Navy cargo ship in a

Brantley and Reist (1994) investigated the exposure to respirable dust at ten ferent blast cleaning sites where quartz sand was used Their results revealed that

dif-in general, downwdif-ind respirable silica concentration varied as distance raised Theconcentration of respirable silica (mg/m3) reduced with distance from the source(feet) according to the following relationship:

The geometry of the worksite and the position of the workers affected trations observed by orders of magnitude The values measured for respirable dustvaried between 0.01 and 10 mg/m3 Randall et al (1998) reported on measurementsperformed during the removal of lead-based paint from a steel bridge with blast

concen-Fig 7.9 Effects of abrasive type on the particle size distribution functions of dust (Kura, 2005)

cleaning The authors measured total dust, respirable dust, total lead exposure andthe exposure of respirable lead Results of these measurements are listed in Table 7.2.The values are all above the permissible limits This situation required the implemen-tation of feasible engineering and work practice controls and the provision of personalprotective equipment (PPE) and hygiene facilities supplemented by use of respirators.Particle size distributions of airborne particles from blast cleaning operationswere analysed by Kura (2005) with different methods Some results are plotted inFig 7.9, and it can be seen that the type of abrasive determined the size distributionfunctions Steel grit formed rather large dust particles, whereas the dust particleswere small for bar shot

Table 7.2 Air sampling analysis results from the removal of paint from a steel bridge (Randall

et al., 1998)

Sampling point Exposure in μ g/m 3

Total dust a Respirable dust b Total lead Respirable lead c

Kura et al (2006) investigated the effects of nozzle pressure, abrasive feed rate(in terms of number of turns of metering valve) and abrasive mass flow rate on theemission of dust during blast cleaning Dust emission increased as the nozzle pressureincreased if painted panels were blast cleaned If rusted panels were blast cleaned, dustemission was almost independent of nozzle pressure The influence of the abrasivefeed rate on the emission of particulate matter was sensitive to the nozzle pressure.

For low and moderate pressures ( p= 0.55–0.69MPa), the emission increased with an

increase in the number of valve turns For higher pressures ( p= 0.83 MPa), however,

the emission showed maximum values at a moderate number of turns The emissionsfor rusted panels were almost independent of the abrasive mass flow rate, whereasthe emissions for painted panels again showed a complex relationship to abrasivemass flow rate and nozzle pressure Results reported by Kjernsmo et al (2003) arepresented in Fig 7.10 The emission of respirable dust increased for higher nozzlepressures It can also be seen that quartz sand generated more dust than copper sand at

equal nozzle pressures ( p= 0.7 MPa) But this trend turned upside down if water was

Fig 7.10 Effects of abrasive type, nozzle pressure and water addition on the formation of

res-pirable dust (Kjernsmo et al., 2003)

added to the nozzle flow; in that case, the dust emission was higher for the copper slag

compared with that of quartz sand at equal high nozzle pressure ( p= 1.0 MPa) and

equal water flow rate (1.1 l/min) The graphs also illustrate the effect of water addition.The version with the highest amount of added water (4.5 l/min) generated the lowestdust level among all tested configurations

Greenburg and Winslow (1932) performed an early thorough study into the fects of location, abrasive type and fresh air supply on the concentration of dustduring blast cleaning operations Some results are listed in Table 7.3 It can be seenthat the use of a mineral abrasive (sand at that time), even when mixed with a metal-lic abrasive material, created much higher dust concentrations compared to the use

ef-of a metallic abrasive Kjernsmo et al (2003) reported on the effects ef-of abrasive type

on respirable dust concentration As shown in Fig 7.11, quartz sand generated thehighest amount of dust (which agreed with the results shown in Fig 7.10), whereascast iron generated very low dust levels Mineral-based abrasive materials are usu-ally more critical to dust formation compared with metallic abrasive materials.Kura (2003) and Kura et al (2006) provided the following statistical model forthe assessment of parameter effects on dust emission during dry blast cleaning:

Ef= a1+ a2· p + a3· ˙mp· a4· p2+ a5· ˙m2

P+ a6· p · ˙mP (7.2)

Here, Efis a specific dust emission factor, given in g/ft2 The pressure is given inpsi and the abrasive mass flow rate is given in lbs/min This relationship holds for

coal slag and bar shot, and for air pressures between p= 0.55 and 0.83 MPa The

constants a1to a6are regression parameters whose values as listed in Table 7.4.Plitzko et al (1998) investigated the effects of abrasive type and water addition

on the concentration of respirable dust during the blast cleaning of metal substrates.Some of their results are plotted in Fig 7.12 It is clear from this graph that eventhe use of a slurry system (method “5”) could not avoid the exposure of impermis-sibly high dust concentrations For dry blast cleaning with quartz, the permissibleworkplace limit was exceeded by a factor of 940 The use of an alternative abrasivematerial and the addition of water allowed for the reduction of this value, but thepermissible limit was still exceeded by a factor of 4

Katsikaris et al (2002) noted an effect of the desired substrate surface ness on the concentration of respirable dust The respirable dust concentration was

cleanli-399μg/m3for a cleanliness degree of Sa 2 and 525μg/m3for a cleanliness degree

of Sa 21/2

Table 7.3 Results of dust measurements for different abrasive materials (Greenburg and

Winslow, 1932)

Abrasive material Dust concentration in 10 6 particles per cubic metre

Minimum Maximum Average

Sand/steel mixture 1.4 66.9 27.8

Fig 7.11 Effects of abrasive material on the formation of dust (Kjernsmo et al., 2003) 1 – cast

iron, 2 – aluminium oxide, 3 – aluminium silicate, 4 – olivine, 5 – quartz sand

Dust concentration, especially in confined spaces, can be reduced due to theutilisation of ventilation systems As shown in Fig 7.13, ventilation could dropdust concentration to very low values Critical parameters were ventilation time andsystem size The longer the ventilation time, the lower was the dust concentration

It was also shown that small ventilation systems can work very efficiently

Blast cleaning operators must usually wear respiratory equipment, combinedwith a separate fresh air supply It was already shown in an early investigation byGreenburg and Winslow (1932) that the amount of air delivered is of fundamentalimportance in determining the degree of protection of respiratory devices Results

of their measurements are provided in Fig 7.14 It can be seen that the dust centration under the helmet reduced with an increase in air supply The graphs alsoillustrate the effects of screens in front of the blaster’s eyes A glass screen notablycontributed to a reduction in dust concentration under the helmet

con-Table 7.4 Regression coefficients for (7.2)

Target parameter in g/ft 2 Coefficients

E f for painted steel 263.73 2.58 −57.17 −0.03 −0.85 0.71

E for rusted steel −206.40 4.13 8.99 −0.01 1.04 −0.24

Fig 7.12 Exceeding of critical dust concentrations for different blast cleaning variations (Plitzko

et al., 1998)

Fig 7.13 Effects of ventilation on dust concentration (Mickelsen and Johnston, 1995)

air supply to helmet in m 3 /min

Fig 7.14 Relationships between air supply to helmet, helmet screen design and dust concentration

in helmets (Greenburg and Winslow, 1932)

An extensive database on site measurements of the exposure of workers to totaldust, respirable dust, total crystalline silica and respirable crystalline silica can befound in Heitbrink’s (1999) report

7.5 Emission of Airborne Metals

7.5.1 Airborne Lead

Because many old coatings contain lead, there is a critical situation as the lead mayenrich the operator’s blood due to breathing the aerosol There are the following twocritical levels:

r Action Level (OSHA AL: 30μg/m3): If an operator works in an area at or abovethat level, the employer must give medical surveillance and training in the haz-ards of working with lead

r Permissible Exposure Limit (OSHA PEL: 50μg/m3): This limit is for the averageamount of lead in the air over an 8-h day

Extensive studies have shown that airborne lead concentration does not depend

on the main lead concentration in coating systems to be removed (NIOSH, 1997);the correlation between these parameters is very weak (correlation = 0.22) It is,therefore, the surface preparation method that determines airborne lead Salome

Fig 7.15 Relationship between lead content in paint and airborne lead exposure for different

abra-sive materials (Salome and Morris, 1996)

and Morris (1996) have shown that the lead content also affected the amount ofatmospheric lead exposure during blast cleaning Some of their results are plot-ted in Fig 7.15 A linear correlation with a correlation coefficient of 0.971 wasfound

Blasters and painters are particularly endangered by lead exposure; this wasverified by a comprehensive medical surveillance programme designed to preventload toxicity in bridge workers, including blasters Some results of these studiesare shown in Fig 7.16, and it can be seen that painters and blasters experiencedthe highest blood lead levels among all job categories Cannon et al (1996) per-formed a blood lead monitoring during steel bridge rehabilitation work where leadcontaining paint was blast cleaned The authors did not find remarkable differences

in blood lead levels for blasters, foremen, operators and painters The only job sification with an average blood lead level greater than a critical level of 20μg/dlwas represented by the foremen These rather low and evenly distributed blood leadlevels were probably due to the fact that the worksite was properly managed withongoing training, reinforcement of personal protection and hygiene practices, in-dustrial hygiene monitoring and frequent medical surveillance Conroy et al (1996)monitored the blood lead levels of bridge workers, including blasters, sweepers,foremen, equipment operators, helpers and supervisors They estimated blood leadlevels were in the range between 26 and 77μg/dl Although the certain value de-pended on the type of work and the work season, they were higher compared to thevalues reported in the study mentioned earlier Several reasons accounted for the

clas-Fig 7.16 Blood lead levels for bridge workers (Maurer et al., 1995)

blood lead elevation: the airborne lead emissions were very high for this project(see Table 7.5), there was lack of water, and the workers did smoke and eat atthe construction site Schulz et al (2005) also reported on unusually high bloodlead levels for workers who were involved in a blast cleaning project, includingabrasive collection, machine maintenance and abrasive distribution Reasons for thehigh levels were inappropriate air delivery, use of wrong filters and very high airtemperatures

A number of measurements of airborne lead exposure to blast cleaning siteswere performed over the years Some results are displayed in Table 7.6 It can beread from the table that dry blast cleaning generates rather high quantities of leadcontaining aerosols Further results of systematic measurements of air samplings in

a containment were reported by Jarrett (2003), who found lead concentrations tween 400 and 6,000μg/m3for blasters Lange (2002) monitored the lead exposure

be-Table 7.5 Personal exposure to airborne lead by job title and activity (Conroy et al., 1996)

Activity Airborne lead concentration in μ g/m 3

Range Median Blasters and sweepers 12–4,401 366

Equipment operators 14–1,400 219

Table 7.6 Measured airborne lead levels for different preparation methods

Object/condition Lead level in μ g/m 3 Reference

Hydroblasting

Galvanised communication towers 1.5–29 Holle (2000)

Structural steel construction 2–12 Dupuy (2001)

Dock side container crane 2.2 a Marshall (2001)

0.79 a,b Marshall (1996)

Slurry blast cleaning

Highway overpass structure 10.4–34.4 Anonymous (1997)

40.1–52.7 d Frenzel (1997) Vacuum blast cleaning

Steel bridge 27–76 c Mickelsen and Johnston (1995) Dry blast cleaning

Steel bridge (blaster) 36–4,401 Conroy et al (1996)

Steel bridge (sweeper) 12–3,548 Conroy et al (1996)

Steel bridge (foreman) 12–3,423 Conroy et al (1996)

Steel bridge (equipment operator) 39–1,900 Conroy et al (1996)

Steel bridge (helper) 22–501 Conroy et al (1996)

Steel bridge (operator) 50–450 a Randall et al (1998)

Petrochemical tank 3.31 a,c Frenzel (1997)

Blast cleaning with pliant media

Offshore oil platform (blaster) 4,990 Miles (2000)

Offshore oil platform (containment) 980 Miles (2000)

Tinklenberg and Doezema (1998) constructed a chart where the exposure to erators was plotted against the content of heavy metals in paints, and they suggested

op-a procedure how to reop-ad criticop-al heop-avy metop-al op-amounts in pop-aints For inorgop-anic rich primers, the following examples were provided To keep operator exposure toless than a permissible exposure limit (50μg/m3), the lead concentration in paintsshould not exceed a value of 280 mg/kg To keep operator exposure to less than anaction level limit (30μg/m3), the lead concentration in paints should not exceed avalue of 170 mg/kg

zinc-Kaufmann and Zielasch (1998) reported about long-term air monitoring duringthe refurbishment of a steel bridge in Switzerland The job was started with dryblast cleaning However, this method was soon replaced by hydroblasting, mainly

Fig 7.17 Long-term air monitoring during steel blasting (Kaufmann and Zielasch, 1998)

because of the high dust emission that exceeded regulatory limits This situation isillustrated in Fig 7.17 Note that during the introductory phase of the project, wheredry blast cleaning was applied, the legal limit of 70μg/m3was exceeded After dryblast cleaning was replaced by a hydroblasting method that featured a robotic tool

as well as limited gun operations, the regulatory limit could be met during the entireproject which lasted over three years (1991–1994)

7.5.2 Other Airborne Metals

A high number of carefully measured data on metal concentrations in airborne ticulates emitted during blast cleaning tests were reported by Kura (2005) for anumber of abrasive materials

par-The exposure of workers to a number of heavy metals during different blastcleaning applications in a ventilated blasting chamber was investigated byTinklenberg and Doezema (1998) Results of their study are listed in Table 7.7 Thedata clearly indicated that during blast cleaning, the operators were exposed verywell above the permissible exposure limits for lead and cadmium Even for a paintwith a reduced level of cadmium and for cured paint, blast cleaning resulted in expo-sure above the permissible exposure limits for both lead and cadmium Tinklenbergand Doezema (1998) designed a chart where the exposure to operators was plottedagainst the content of heavy metals in paints, and they suggested a procedure how toread critical heavy metal amounts in paints To keep operator exposure to less than

Table 7.7 Exposure to heavy metals during dry blast cleaning of zinc-rich coatings (Tinklenberg

and Doezema, 1998); based on 2 h of sampling

Materials Metal content in μ g/m 3

Arsenic Zinc Lead Cadmium Chromium Copper

∗Results below detectable limits

a permissible exposure limit, the cadmium concentration in an inorganic zinc-richprimer should – as an example – not exceed a value of 12 mg/kg Airborne cad-mium was measured during a bridge blast cleaning project by Conroy et al (1996).The median concentration inside the bridge containment was 15.7μg/m3, whichexceeded the permissible level (OSHA PEL: 5μg/m3)

Results of an exhausting study which considered surface condition and wind locity are provided in Table 7.8 A general trend between wind velocity and metalemission could not be installed based on this data Iron was the dominating metal

ve-in all cases, even where a pave-inted surface was blast cleaned Permissible exposuretimes depended very much on the location of the blast cleaning application Someexamples are listed in Table 7.9 It can be seen that the permissible limit for arsenicexposure was reached after 15 min in a blasting room, whereas it was reached after

60 min for an outdoor application

Table 7.8 Summary of emission factors for metals blast cleaned with silica sand (Kinsey

et al., 1994; Anonymous, 1997)

Operation

conditions

Emission factor in kg per kg abrasive

Cadmium Chromium Iron Manganese Nickel Lead Clean surface