The European Organisation for the Safety of Air Navigation

EUROCONTROL – Univ. Politehnica of Bucharest & ROMATSA Project Volcanic Ash Safety

presented at ICAO IVATF, Montreal, 11-15 July 2011

Ash Safety

Prof. dr. ing. C. BERBENTES.l. dr. ing. mat. A. BOGOIConf. dr. ing. T. CHELARUS.l. drd. Ing. C. E. CONSTANTINESCU, MBAProf. dr. ing. S. DANAILAD. DIMITRIU, PhD

Drd. Ing. V. DRAGANS.l. dr. ing. D. D. ISVORANUE. HALIC, M.D.S.l. dr. ing. L. MORARUConf. dr. dr. ing. O. T. PLETER, MBA

As. drd. A. RAJNOVEANU, M.D.Prof. dr. ing. V. STANCIUConf. dr. ing. M. STOIA-DJESKADrd. Ing. D. C. TONCUR. ULMEANU, M.D. PhD

Studies on the Measurement and the Effectsof the Volcanic Origin Particles in Suspension in the Atmosphere

on the Safety of Aircraft

Ash Safety

Presentation Title Ash Safety Research Results

Version / Date / Author 1.1 / 1.07.2011 / OTP

Beneficiary EUROCONTROL

Keywords volcanic ash, volcanic dust, sand aerosols, impact on aviation, impact on flow management

4

Presentation Overview

1.Introduction – research subject, objectives, and relevance

2.Risks, vulnerability, and the scale of the phenomenon

3.Ash vs. Dust; sand aerosols; why does particle size matter?

4.Extent of the danger area: volcanic ash cloud estimation

5.Human health effects

6.Concentration measurement and forecasting

7.Selected conclusions

5

• Research Subject„Studies on the Measurement and the Effects of the Volcanic Origin Particles in Suspension in the Atmosphere on the Safety of Aircraft”

• Research Objectives: to provide objective, relevant, and scientifically validated information for the future decisions on the air traffic in volcanic ash contaminated atmosphere, for the best trade-off for the most fluent air traffic under the circumstances, with uncompromised flight safety.

Ash Safety Project was triggered by Eyjafjalla

6

Read the Scientific Report to find that …

• Visible volcanic ash cloud is dangerous to aviation, but it does not travel very far (1-200 NM, max. 550 NM down the wind in the greatest eruption in history, Pinatubo 1991);

• Volcanic dust contamination in concentrations comparable to that current for sand aerosols (4∙10─3 g/m3) is not a safety issue and rather a maintenance issue (like flying at Ryadh or Cairo);

• Tactical avoidance of volcanic dust contamination based on on-board instruments is not practical;

7

• Worries about volcanic ash/dust contamination impact on the human respiratory system of the occupants of an aircraft are not justified;

• Volcanic Ash/Dust Concentration is a noisy random variable, very difficult to measure with certainty and consistency; the concentration measurements should be averaged on a cubic hectometre scale (1,000,000 m3; the hectometric principle);

• Concentration measuring unit: 1 kg/hm3 = 10−3 g/m3

Read the Scientific Report to find that … (2)

8

• Volcanic Dust Concentration is less important than the size distribution of particles; volcanic dust particles resulting from differential sedimentation in the natural atmospheric dispersion process do not cause abrasion and/or glass deposits inside the turbofan engine, regardless the concentration;

• The best option to address volcanic ash/dust problems in the future is a quick reaction algorithm to estimate volcanic ash cloud and a prediction based on an Euler dispersion model with data assimilation from hectometric in-situ sensors

Read the Scientific Report to find that … (3)

9

Parts / Occupants Cause Effect ResponseTurbine engines fuel injection and combustor

deposits of melted ash (glassy coatings)

surge, shut-down, difficult restart in flight

idle thrust, evasive maneuver

Turbine engines clogging the turbine cooling vents

overheating idle thrust, evasive maneuver

Pitot-static clogging the sensors unreliable air speed indications

attitude-based flying, indicated air speed deducted from ground speed and wind velocity

Turbine engines abrasion with hard particles wear of fan, compressor, turbine, transmission

idle thrust, evasive maneuver

Pneumatic controls clogging the vents failure evasive maneuver

Windshield, body, wings, empennage

cracks, abrasion with hard particles

wear, opaqueness evasive maneuver

Avionics, on-board instruments

clogging air-cooling vents, electrostatic discharges

overheating, malfunction evasive maneuver

Human occupants breathing contaminated air, eye cornea contact with ash/dust particles

respiratory problems, eye damage

nose breathing, replace contact lenses with eyeglasses

Turbine engines, body and instruments metallic parts

acidity, exposure to associated SO2 and sulfurous acid

corrosion (in time) maintenance check and replacement

What can go wrong in a VA encounter

10

Air Breathing Order of Magnitude

Description Affected Hardware or Liveware

1,000 m3/s High flow non-filtered air breathing

turbine engines

100 m3/s Directly exposed to airflow

windshield, empennage, body and wing

0.01 m3/s Low flow non-filtered air breathing

human occupants, Pitot-static sensors, computers, electrical engines and other air-cooled parts

Irrelevant Air breathing through filters

piston engines, air-cooled parts with air filters

Vulnerability ~ Air Breathing Flow

11

Turbofan engines are huge vacuum cleaners, sucking an average of 1,000,000 m3 = 1 hm3 each in 10 minutes of flight

The Silica particles in the core flow will be deposited as glass in the combustion chamber and on the HPT

One kilogram of deposits is enough to cause turbine overheating and even engine failure (restarting is possible though outside the contaminated area)

Vulnerability is Critical for Turbofan Engines

12

Characteristic time scope = 10 minutes of flight = exposure of an average turbine engine to 1,000,000 m3 = 1 hm3 of air

10 minutes is the maximum exposure of an aircraft engaged in an evasive manoeuver after an unanticipated volcanic ash encounter

Scale of phenomenon = 1 Cubic hectometre

13

Example of quantity of contaminant accumulated in 10 minutes of flight in a

concentration of 2∙10─3 g/m3

Air Breathing Order of Magnitude

Quantity of contaminant accumulated in 10 minutes flight at a concentration of 2 10∙ ─3 g/m3

Affected Hardware or Liveware

1,000 m3/s 1.2 kg turbine engines100 m3/s 120 g windshield, empennage, body

and wing0.01 m3/s 12 mg human occupants, Pitot-static

sensors, computers, electrical engines and other air-cooled parts

Irrelevant 0 g piston engines, air-cooled parts with air filters

Terminology discriminator:

Volcanic Ash = 1/16 mm – 2 mm(Coarse ash)

Volcanic Dust < 1/16 mm(Fine ash)

Volcanic Ash Volcanic Dust

Particle size (µm) 63 - 2000 < 63

Location 1-200 NM around the volcano

Floats around the world

Sedimentation time ≈ 1/2 hour (for 1 mm) 23 Days (for 10µm)

14

15

Volcanic Dust < 1/16 mm(Fine ash)

Volcanic Ash = 1/16 – 2 mm(Coarse ash)

Ash vs. Dust

16

Volcanic Dust Haze(Contamination)

Thin layers of dust only visible from selected viewing angles or from a far distance e.g. satellite

Volcanic Ash Cloud

Cloud clearly visible to naked eye from all angles, clear boundary

Visible Volcanic Ash Cloud vs. Volcanic Dust Contamination

17

Eruption Case Study: H=10 km, W=50 kts

Falling speed (m/s) from:

1 mm (ash) 100 m (ash)

10 m (dust)

1 m (dust)

10,000 m 6.0 0.8 0.005 8·10−5

8,000 m 5.7 0.7 0.005 8·10−5

6,000 m 5.5 0.7 0.005 7·10−5

4,000 m 5.3 0.7 0.004 7·10−5

2,000 m 5.1 0.6 0.004 7·10−5

Average falling speed (m/s)

5.5 0.7 0.005 7·10−5

Ash: Dust

1/16 mm = 62.5 mm

Ash: visible,dangerous,local / short term

Dust: Less threatening,

globe trotter /long term

1 mm (ash) 100 m (ash) 10 m (dust)

1 m (dust)

Number of particles in 1 m3 at a concentration of 2·10−3 g/m3 (2 kg/hm3)

3 2,728 2,728,370 2,728,370,453

18

1 mm (ash)

100 m (ash)

10 m (dust)

1 m (dust)

Sedimentation Time (h) 0.5 4.0 555.6 39,682.5Distance travelled (NM) 25 200 27,780 1,984,125

Sand Blasting Effect on Particle Size

Effects of particle impact velocity and particle size on substrate roughness, from Weidenhaupt, W., 1970, Über den Einfluß der Entzunderungs- und Ziehbedingungen auf die, Oberflächen-Feingestalt von Stab- und Profilstahl, Dissertation, RWTH Aachen, Germany. as referenced by Momber, A., 2008, Blast Cleaning Technology, Springer

20

Abrasion Caused by Ash vs. Dust

Ash follows airflow streamlines imperfectly and impacts walls: inertial forces > aerodynamic forces

Dust follows airflow streamlines: inertial forces << aerodynamic forces

21

Glass Coating / Clogging by Ash vs. Dust

22

Relative Safety Risk vs. Size: 0.1-1000 m

(qualitative representation)

23

Highest-Risk Dust Particle Segment: 1-10 m

Turbofan engines (tests on simulated engines):Larger> The turbofan functions like a centrifugal separator

(no issue with the volcanic ash/dust in the by-pass flow)Smaller> Particles exit the engine without touching any wall

Human respiratory system:Larger> filtered by body barriersSmaller> get out with expiration(not retained in the lungs)

Volcanic Ash / Dust / Sand Aerosols

Volcanic Ash Cloud Volcanic Dust Contamination

Sand Aerosol Contamination

Visibility Clearly visible from all angles, easily identifiable by dark colour, distinct boundary

Visible only from selected angles, hard to distinguish, visible in satellite imagery

Visible sometimes from selected angles, visible in satellite imagery

Where? Within 1-200 NM of the eruption up to 500 NM

Very large areas (>1000 NM in size)

Large areas

Typical Concentrations(1 kg/hm3 = 10−3 g/m3)

1000 kg/hm3 1-100 kg/hm3 1-100 kg/hm3

Particle size range (µm)

1-2000 1-40 1-50

Floatability in atmosphere (age)

12 Hours (due to ash-dust differentiated sedimentation)

6 Days (traces remain for years)

3 Days

24

Volcanic Ash Cloud

Volcanic Dust Contamination

Sand Aerosol Contamination

Aviation Safety Risk

Serious incidents, no injury accidents

None on record None

Impact on aviation

Local Global due to misinterpretation

Maintenance issues

Source: Prof. Ulrich Schumann, DLR Institute of Atmospheric Physics, NASA Earth Observatory

25

26

Severity vs. Frequency Safety Risk

16/02/2007 = 14 aircraft: windshield failures within 3½h19/01/2011 = 3 aircraft: windshield failures within ½h

27

Ash-Sized Sand – a New Safety Concern (Denver, CO)

Source for photos: www.airliners.net

Skywest Canadair CRJ-700 on behalf of United Airlines, flight OO-6761/UA-6761 from Denver, CO to Cleveland, OH (USA), reported a cracked windshield during initial climb out of runway 34R and returned to

Denver .

Skywest Canadair CRJ-200 on behalf of United Airlines, registration N978SW performing flight OO-6576/UA-6576 from Denver, CO (USA) to Winnipeg, MB (Canada), reported a windshield cracked during initial climb out of runway 34L and returned to Denver

Delta Airlines Boeing 757-200, flight DL-1916 from Denver, CO to Atlanta, GA (USA), during initial climb out of runway 34L reported a cracked windshield and returned to Denver

28

Future Eruption First Reaction Checklist

Location of the eruption / time: LAT, LONG, HHMMz, DDMMYY

{repeat until eruption ends}

How tall is the eruption column? ECH (m AMSL)

Download wind profile in the area (e.g. from NOAA): WD/WV

Calculate how far will the volcanic ash cloud go: VAMAX

Draw a contour with VAMAX as major axis on the map: DA

Ash

4D

29

Volcanic Ash Danger Area

Shape: defined by the variability of wind direction and amplitude of wind velocity

30

Volcanic Ash Danger Area

How far does the VA travel?

Function of: H (height of eruption column), FL (flight level), and WV (wind velocity)

Eruption column H (m) 30000Flight Level (100 ft) 100Wind Velocity (kts) 50Danger area (NM) 535

Pinatubo 1991Eruption column H (m) 9000Flight Level (100 ft) 100Wind Velocity (kts) 100Danger area (NM) 236

Eyjafjalla 2010

Eruption column H (m) 3500Flight Level (100 ft) 30Wind Velocity (kts) 20Danger area (NM) 21

Etna 2011

31

Why 500 NM for Pinatubo = 250 NM for Eyjafjalla?

Slide from Jacques Renvier, CFM/SNECMA, Atlantic Conference on Eyjafjallajokull, Keflavik 2010

32

Why 500 NM for Pinatubo = 250 NM for Eyjafjalla?

(Kilauea, Hawaii

28 years continuous eruption)

33

Conclusions on Human Health Effects

What was studied: passengers (short-term) and crew (long-term) exposure to volcanic dust concentrations of 4∙10─3 g/m3

“Well under limits for short time exposure and even lower than those for chronic exposure”

“Reasonable to anticipate no risk for silicosis or lung cancer in passengers and crew members”

“Concentration lower than environmental Silica aerosols in some parts of the world (e.g. Ryadh, Cairo)”

34

Concentration: volatile, noisy, measured indirectly

Direct and consistent methods to measure concentration in real time: none

Indirect methods, subject to large methodical errors and measuring historical values of concentration along a line of sight/movement:

• In-situ sampling;• Ground up-looking LIDAR;• Airborne down-looking LIDAR;• Ground optical photometre;• Satellite infrared image.

35

Concentration measurement problems: 1. scale of phenomenon and 2. consistency

All these technologies deployed in the same area would yield different results

36

Hectometric Principle: Why do concentration measurements require averaging at the 1 hm3 scale?

In this example we assume we want to measure the brightness of a halftone image

37

Future sensors are shortsighted and the scale of the phenomenon is larger

38

Types of concentrations Actual Measured Forecasted Blindly (open-

loop)

Forecasted with Data

AssimilationWhen they are available (hours) never T+2H T−18H (up to

T−180H)T−18H (up to T−180H)

Uncertainty due to initial data of the eruption (orders of magnitude)

- - 2 - 3 0.2 - 1

Uncertainty due to the Eulerian diffusion model (orders of magnitude)

- - +1 every 24 hours

0.1 - 0.4

Uncertainty due to the measurement techniques

- 0.1 – 1 - -

Area coverage - Very local1 Global GlobalOverall errors 0 Significant Very large and

rapidly increasing in time

Large

Relevance to IFR flight operations - Tactical avoidance

Flight planning Flight planning

Relevance to ATM - Forecasts Validation

Airspace management, Flow management

Airspace management, Flow management

[1] Except the satellite images, which provide a global view

Airborne in‐situ hectometric concentration measurement unit

isokinetic

39

VADHCMUVolcanic Ash/Dust Hectometric Concentration

Measurement Unit

1. Airborne 6 channel (3 wavelengths plus polarizations) backscattering infrared LIDAR,

2. Ground based (mobile) optical photometre, and

3. Airborne in situ hectometric concentration ‐measurement unit.

A consistent and systematic measurement technology for data assimilation:

40

41

Selected Conclusions

• Extending the term of volcanic ash cloud as per ICAO Doc9691 to the volcanic dust haze is wrong and misleading;

• Size of the particles matters more than concentration as impact on aircraft and aircraft turbofan engines;

• Predicting concentrations using a dispersion model with data assimilation from a systematic daily measurement scheme using hectometric in-situ samplers is the best choice for a way forward;

• Immediate reaction to a VA threat should be based on a simple kinematic model of the extension of the volcanic ash cloud and not on chasing the dust going round the globe.

Please, read the final report for more!