Storage Tank Assessment

Matrix Inspection & Engineering Inc.

A Most Difficult Application

Early work focussed on leak detection

Hartford Steam Boiler

Large number of tests, some notable successes

DnV (personnel now with Matrix)

Better technology, much effort to develop an understanding of tank acoustics and noise.

API

Confirmed earlier work, especially problems with noise. Leak detection and (especially) location not reliable enough for widespread use.

What can AE do for Tanks?

Qualitative assessment of tank (floor) condition

AE activity from corrosion process, scale fracture and leaks. Experimental work shows scale build-up is a significant factor.

Simple approach to a complex problem: more AE indicates more damage. Similar to well established procedures for FRP and metal vessels.

Published work by Shell (Netherlands) shows the method has value.

How Reliable is AE Tank Testing?

Study published by Shell (Amsterdam) in 1998

Down to earth facts and figures from a large population of tests in Europe and the Middle East.

AE results correlated with inspection findings.

AE Rating Vs Internal Inspection

What Does This Study Tell Us?

Putting the Technique to Work

Over-calls may be reduced with proper attention to noise, filtering and evaluation.

Results improved by regular testing and historical data.

The accuracy of AE results must be compared with conventional methods for selecting tanks for inspection. How many ‘good’ tanks do you inspect?

Problems With Noise

Extraneous noise and complex acoustics make interpretation difficult - leads to over-calls

Need for low frequency sensors and high sensitivity.

Roof seal movement and other noise sources.

Acoustic signals, including noise, reflect and reverberate around the inside of the tank.

No controlled loading to correlate with AE activity.

Dealing With Noise

Treat a tank as a three dimensional problem - the floor is the area of interest but not the only source of AE activity.

Advanced filtering techniques, ‘guard’ sensors and 3-D plotting.

Test when conditions are right.† Matrix uses a portable weather station to monitor wind speed.

Use sensors with good case isolation. Digital signal processing for best sensitivity and measurement.

Under-calls: Damage not Detected

Insufficient sensitivity

Heavy oil products, coatings and sludge attenuate AE signals (compared to clean, product tanks).

Excessive noise may mask ‘real’ AE signals.

Some problems with linings (in bad condition).

Short Monitoring Time

AE data collected over 1 hour, some corrosion sites may not be active over that time span.

Complex acoustics

One or two dominant AE sources may mask weaker sources of AE activity.

Improving Test Effectiveness

Use Good Testing Practice

Adjust evaluation criteria for tank type - more severe for large crude tanks than small product tanks.

Do not test if background noise is excessive.

Increase monitoring time for some tanks.

AE as Part of a Program

Incorporate other data (RBI) and measurements (Lorus, ultrasonic scans of roof and shell).

Repeat testing to monitor trends.

Location of AE Sources

Location of AE sources in tanks is very hit and miss. Multipath (reflected) signals and multiple sources of AE activity can cause large errors.

Location techniques are important for separating clusters of AE activity for evaluation purposes but the actual location of AE sources is uncertain.

If AE data shows a tank has damage, plan to inspect the whole floor.

Test Procedure - Preparation

Fill to a minimum of 50%.

Block in all lines, including roof drains.

12-24 hour settling time after filling.

Prepare permits for access and use of electrical equipment.

Supply 110V power.

Test Procedure - Measurement

5-12 primary sensors and 3-6 guard sensors depending upon tank diameter.

Check sensitivity, measure speed of sound.

Identify and deal with sources of noise. Monitor wind speed continuously.

Monitor and record AE activity (1-2 hr).

Evaluate data.

Evaluate AE Data - Filtering

Identify and remove noise

Evaluate AE Data - Plotting

Location plots show the distribution of AE activity.

Need More than a Plan View

AE Versus Internal Inspection

AE Testing

Test in one day or less depending upon tank size.

Tank still for 12-24 hours.

Low cost, little preparation needed.

Indication of damage, not definitive.

Cost effective way to prioritize tanks for internal inspection and postpone inspection of ‘good’ tanks.

Internal Inspection

Days to weeks depending on cleaning and prep.

Tank empty and out of service for weeks or months.

Cost $50k to >$150k before inspection begins.

Full documentation of defects and metal loss.

Can only inspect a few tanks per year. Important to pick the right ones.


Early work focussed on leak detection
	Hartford Steam Boiler
		Large number of tests, some notable successes
	DnV (personnel now with Matrix)
		Better technology, much effort to develop an understanding of tank acoustics and noise.
	API
		Confirmed earlier work, especially problems with noise. Leak detection and (especially) location not reliable enough for widespread use.


	Qualitative assessment of tank (floor) condition
		AE activity from corrosion process, scale fracture and leaks. Experimental work shows scale build-up is a significant factor.
		Simple approach to a complex problem: more AE indicates more damage. Similar to well established procedures for FRP and metal vessels.
		Published work by Shell (Netherlands) shows the method has value.


	Study published by Shell (Amsterdam) in 1998
		Down to earth facts and figures from a large population of tests in Europe and the Middle East.
		AE results correlated with inspection findings.


		Over-calls may be reduced with proper attention to noise, filtering and evaluation.
		Results improved by regular testing and historical data.
		The accuracy of AE results must be compared with conventional methods for selecting tanks for inspection. How many ‘good’ tanks do you inspect?


	Extraneous noise and complex acoustics make interpretation difficult - leads to over-calls
		Need for low frequency sensors and high sensitivity.
		Roof seal movement and other noise sources.
		Acoustic signals, including noise, reflect and reverberate around the inside of the tank.
		No controlled loading to correlate with AE activity.


		Treat a tank as a three dimensional problem - the floor is the area of interest but not the only source of AE activity.
		Advanced filtering techniques, ‘guard’ sensors and 3-D plotting.
		Test when conditions are right.† Matrix uses a portable weather station to monitor wind speed.
		Use sensors with good case isolation. Digital signal processing for best sensitivity and measurement.


	Insufficient sensitivity
		Heavy oil products, coatings and sludge attenuate AE signals (compared to clean, product tanks).
		Excessive noise may mask ‘real’ AE signals.
		Some problems with linings (in bad condition).
	Short Monitoring Time
		AE data collected over 1 hour, some corrosion sites may not be active over that time span.
	Complex acoustics
		One or two dominant AE sources may mask weaker sources of AE activity.


	Use Good Testing Practice
		Adjust evaluation criteria for tank type - more severe for large crude tanks than small product tanks.
		Do not test if background noise is excessive.
		Increase monitoring time for some tanks.
	AE as Part of a Program
		Incorporate other data (RBI) and measurements (Lorus, ultrasonic scans of roof and shell).
		Repeat testing to monitor trends.


	Location of AE sources in tanks is very hit and miss. Multipath (reflected) signals and multiple sources of AE activity can cause large errors.
	Location techniques are important for separating clusters of AE activity for evaluation purposes but the actual location of AE sources is uncertain.
	If AE data shows a tank has damage, plan to inspect the whole floor.


	Fill to a minimum of 50%.
	Block in all lines, including roof drains.
	12-24 hour settling time after filling.
	Prepare permits for access and use of electrical equipment.
	Supply 110V power.


	5-12 primary sensors and 3-6 guard sensors depending upon tank diameter.
	Check sensitivity, measure speed of sound.
	Identify and deal with sources of noise. Monitor wind speed continuously.
	Monitor and record AE activity (1-2 hr).
	Evaluate data.


	AE Testing
		Test in one day or less depending upon tank size.
		Tank still for 12-24 hours.
		Low cost, little preparation needed.
		Indication of damage, not definitive.
		Cost effective way to prioritize tanks for internal inspection and postpone inspection of ‘good’ tanks.
	Internal Inspection
		Days to weeks depending on cleaning and prep.
		Tank empty and out of service for weeks or months.
		Cost $50k to >$150k before inspection begins.
		Full documentation of defects and metal loss.
		Can only inspect a few tanks per year. Important to pick the right ones.