NATIONAL TRANSPORTATION SAFETY BOARD
PIPELINE SAFETY HEARING
Inspection and Integrity Verification
Conference Center and Board Room
National Transportation Safety Board
Washington, D.C.
Wednesday, November 15, 2000
9:30 a.m.
Day 1 Transcript
Day 2 Transcript
Board Members Present:
JIM HALL, Chairman
CAROL CARMODY
JOHN J. GOGLIA
JOHN HAMMERSCHMIDT
GEORGE W. BLACK, JR.
Board Staff Present:
BOB CHIPKEVICH
JOSEPH KRIS
Hearing Officer
CLIFF ZIMMERMAN
Pipeline Accident Investigator and IIC
ROD DYCK, Associate Director
Pipeline Division
JIM WILDEY, Chief
Materials Laboratory
Office of Research Engineering
A G E N D A
AGENDA ITEM: PAGE:
Opening Statement 3
Chairman Jim Hall
National Transportation Safety Board
Honorable James Oberstar 9
8th Congressional District
State of Minnesota
Overview 37
National Transportation Safety Board
Staff
Remarks 24
Honorable Kelley S. Coyner, Administrator
Research and Special Programs Administration
Stacey Gerard, Associate Administrator
Pipeline Safety
Panels:
In-Line Inspection Service 49
Tom Sawyer, PII North America, Inc.
Ravi Krishnamurthy, PII North America, Inc.
John Parsons, Tuboscope
Afternoon Session
Integrity Assessment 123
H. Noel Duckworth, Consultant
Dr. John F. Kiefner, Kiefner & Associates
Pipeline Operators 168
Vic Yarborough, Colonial Pipeline Company
Rich Turley, PE, Marathon Ashland PipeLine
LLC
J. Andrew Drake, PE, Duke Energy
Elden R. Johnson, Alyeska Pipeline Service Company
Researchers 235
Dr. Tom Bubenik, Battelle
Dr. Brian Leis, Battelle
Al Crouch, Southwest Research Institute
P R O C E E D I N G S
9:36 a.m.
CHAIRMAN HALL: Before I gavel this hearing to order, I would like to make a brief announcement of welcome to our guests to the National Transportation Safety Board. You are -- we're very pleased to welcome you to our new Board Room.
There is an important announcement I would like to make in regard to any possible emergency evacuation of this room. In the event of an emergency, such as a fire, the building alarm system will activate, and a voice message will instruct persons to vacate the building.
You should proceed to the nearest exit. There are emergency exits up front, to the left and to the right of this platform and in the rear of the room.
Also, for your convenience, restrooms and telephones are located in the foyer on the left as you exit the room.
If there is anything that either myself or the other members of the National Transportation Safety Board can do to assist you while you are visiting with us, please do not hesitate to ask.
I'd like to open this Pipeline SafetyHearing, and I would like to note the excellent attendance that we have this morning, and I am very grateful for that, as we attempt to put attention on an extremely important safety issue that does not receive as much attention as many of the other modes of transportation that the Board is involved in.
Because of the importance the Board places on this critical safety issue, all five Board Members will participate in this hearing over the next two days. The safe transportation of natural gas and liquid petroleum products is vital to meeting the energy needs of every community in our country. Pipelines today provide a vital transportation service to America.
Over 2.1 million miles of pipelines criss-cross our great country, many of them running under our cities, our towns, our neighborhoods, and our playgrounds.
Last year, pipelines delivered over 13 billion barrels, 558 billion gallons, of petroleum products, such as crude oil, gasoline, diesel fuel, and heating oil, to customers all across the nation.
In addition, the number of customers using natural gas in this country has exceeded 60 million, and approximately 70 percent of new homes now havenatural gas service.
As many of you know, the National Transportation Safety Board has been the eyes and ears of the American people at accident sites for over three decades, and in those 30 years, we have seen too many tragedies or near tragedies that were all caused by the same fundamental problems.
We have scheduled this two-day Pipeline Safety Hearing to address two of those problems. Today, we will focus on Pipeline Inspection and Integrity Verification. Tomorrow, we will examine Leak Detection and Response.
The Safety Board is currently investigating six pipeline accidents that have occurred since last year, where time-dependent defects are being examined. Those accidents occurred in Knoxville, Tennessee, Bellingham, Washington, in which three young men were tragically killed, Winchester, Kentucky, Greenville, Texas, Chalk Point, Maryland, and Carlsbad, New Mexico, in which 12 people died.
Many of the hazardous liquid and natural gas transmission pipelines in our country are 30 to 50 years old. The Liquid Products Pipeline that ruptured in Bellingham, Washington, on June 10th, 1999, was constructed in 1966. The Natural Gas Pipeline thatruptured near Carlsbad, New Mexico, on August 19th, 2000, was constructed in the early 1950s.
Although age alone does not indicate that a pipeline is unsafe, it does make determining the integrity of pipelines increasingly important.
13 years ago, the Safety Board recommended that the Research and Special Programs Administration, referred to in this town as RSPA, require pipeline operators to periodically inspect their pipelines to identify time-dependent defects that may prohibit safe operations.
Just last week, RSPA issued a Final Rule to require pipeline operators who operate 500 or more miles of hazardous liquid pipeline to establish an integrity management program for high-consequence areas.
According to RSPA, this rule will cover 87 percent of all hazardous liquid pipelines. We will be examining this Final Rule closely in the next two days, but it appears to be the first step to ensuring that pipelines are properly inspected and tested.
The lack of timely recognition of a pipeline rupture is another recurring problem we see repeatedly in our investigations. Following a pipeline rupture, controllers often continue to operate a pipeline orrestart a system that had shut down rather than promptly shutting the system down and isolating the leak. This failure to recognize a problem in a timely manner can significantly add to the accident severity.
Over the next two days, we will be examining the technology available to address these important safety issues, looking at the limitations of the current technology and actions that are needed to identify time-dependent pipeline defects before they reach critical size and to recognize promptly when a failure does occur.
Before I begin, I want to thank each of the panelists who have agreed to participate in our hearing. I also want to thank Tuboscope and PII North America for bringing their in-line inspection tools to our hearing, so that all of us can better understand how they work.
The Tuboscope device is a model of an in-line tool or smart pig. It is located in the lobby just outside the Conference Center. The PII tool will be arriving later this morning. It is used to inspect 12-inch diameter pipelines.
I'm told that it is about 17 feet long and weighs approximately 1200 pounds. So, rather than bring it into this building, we'll display it on theplaza on the floor above us. You'll be able to see it just outside the building at the top of the escalator.
I hope that you'll take the opportunity to examine both of these items in the next few days.
Just a word of caution. The PII tool has a strong magnetic field. Therefore, if you have a pace-maker, you should not get close to the device.
Now, before we begin with our scheduled witnesses, it gives me great pleasure to introduce one of the truly exceptional individuals in the United States Congress today, and a true leader in the field of transportation safety.
Congressman James Oberstar has worked on Capitol Hill since 1963. 26 years ago, he was elected as the representative of the 8th Congressional District of Minnesota, and I'm happy to say he was re-elected again last week without needing a recount.
From 1989 to 1995, he chaired the Aviation Subcommittee. Currently, he is the Ranking Member of the Transportation and Infrastructure Committee, and he serves as an ex-officio member of the Subcommittees on Aviation, Coast Guard and Maritime Transportation, Public Building and Economic Development, Railroads, Surface Transportation, and Water Resources and Environment.
Over the years, Congressman Oberstar has worked ceaselessly to improve the safety of the nation's transportation system. Last month, Congressman Oberstar introduced the Pipeline Safety Act of 2000, legislation to improve the safe operation of hazardous liquid and natural gas pipelines, each kind of pipeline involved in the Bellingham, Washington, and Carlsbad, New Mexico, tragedies.
Congressman Oberstar has been a friend and supporter of the NTSB and its mission, and his presence here today is further testament of his commitment to improve the safety of our nation's pipeline infrastructure.
Ladies and gentlemen, I welcome to the podium Congressman Jim Oberstar.
(Applause)
CONGRESSMAN OBERSTAR: It's on? There it is.
Thank you very much, Mr. Chairman, for that very warm introduction, and, no, we don't need a recount in my district. I got 73 percent of the vote.
In fact, the second-highest vote total in the whole country.
More importantly, I thank you and the members of the Board for again pursuing the role that the NTSB serves so exceedingly well as the indisputableauthoritative, objective guardian of safety in transportation.
The recommendations of this Board on all forms of transportation over the years of its existence, since 1969, have saved countless lives, made transportation safer, made America a better place, but your continued vigilance is necessary because safety is only a matter of seconds. It is a matter of measuring risk, and we can never measure it too carefully or too closely.
I thank you for continuing this extraordinary service of the Board.
On Tuesday, July 8, 1986, a quiet neighborhood in Moundsview, Minnesota, was roused from its slumber when a wall of fire roared down that street. A mother and her six-year old daughter stepped out the front door shocked by the noise, frightened obviously, opened their door and were incinerated.
Mailboxes melted. Trees wilted. The road buckled. A third woman was severely injured. Quarter of a million dollars in property damage was caused. The origin of the fire, a hazardous liquid pipeline running through this neighborhood.
The neighborhood had overgrown the pipeline. It had once been a rural area, now suburban, and in theensuing investigation, it was found that cathodic protection on that pipeline had failed and had gone undetected, and the pipeline had rusted through, and the unleaded gasoline going through that pipeline leaked over an extensive period of time, no one is really sure how long, until it reached a critical mass in that street area, and the fumes rose to the surface as an automobile was driving along that street. A loose exhaust pipe caused a spark, and the road exploded.
I was Chair of the Investigations and Oversight Subcommittee of the House Public Works Committee at the time. We were preparing hearings on the condition of the nation's pipelines. We intensified our work, Mr. Chairman, when that tragedy occurred, and held a hearing in 1987.
We asked the General Accounting Office, in preparation for the hearing, to assess its evaluation -- to reassess its evaluation of OPS, Office of Pipeline Safety, operations. It was clear that OPS did not have sufficient manpower to carry out inspections, to carry them out at appropriate intervals.
The current regulations were not adequate. The federal/state partnership in pipeline inspections was not working. There were factors, such as one callsystems, automatic shut-off valves, clear path, other technologies for response to tragedies, that were grossly inadequate or non-existent. That was 1987.
Our subcommittee made recommendations for the legislative committees to act. They did act to increase the number of inspectors, but an unwilling Administration would not go further with more requirements in either the states or the Office of Pipeline Safety.
In the decade since then, more than 2500 accidents have occurred on the nation's pipelines. We concluded the last session of Congress or virtually have, appropriations aren't completed, but on this issue, there was no agreement on a pipeline safety bill.
We passed -- we were at the point of moving a bill in our committee when the Senate moved theirs, and the advocates for a weaker bill said, "Don't let the perfect be the enemy of the good." Well, that phrase was first uttered by Voltare. "Ne pas laissez le parfait etre l'ennemi de bon". Don't let the good be the enemy of the better.
But Voltare went further and said, "A la vivant, il faut. A la mort, il faut la vrai". To the living, we owe respect. To the dead, we owe the truth. The dead in this case are the innocent victims of pipeline tragedies, and the most recent being those in Bellingham, Washington, Carlsbad, New Mexico, as you've cited, Mr. Chairman.
These are not just numbers, these tragedies, these statistics. These are human beings, whose lives are wrenched, ripped apart, torn asunder. Marlene Robinson, the mother of one of the victims of Bellingham, told a group of members of Congress about her son. He had just graduated from high school. He didn't go off on a party with his friends. He didn't go drinking or carousing. He took his fly rod and went fishing.
While he was fishing, celebrating his graduation, a wall of fumes roared down that river and succumbed him and then exploded and burned three others -- two others. Excuse me. The cause, a ruptured pipeline, gasoline pouring into the creek, fumes roaring ahead of it. "A la mort, il faut la vrai". To the dead we owe the truth, and the truth is that we can do better, and we have to do better on the nation's --monitoring the nation's pipelines.
We have 2.2 million miles of pipelines, carry 617 million ton miles of oil and refined products, 20 trillion cubic feet of natural gas every year, and itcontinues to grow.
Pipeline mileage has grown 10 percent in the last 10 years, but it's growing at the very same time that the nation's suburbanization continues to bring more families near more pipelines.
But as the industry has grown, regrettably, our hearings 13 years ago, this Board's recommendations and investigations, General Accounting Office, congressional committee hearings have shown that safety has declined.
In the last 10 years, the decade of the '90s, there were 2241 major pipeline accidents resulting in death, serious injury or substantial property damage. Those explosions killed 226 people, caused $700 million of property damage and damage to environment, and the General Accounting Office reports that the rate of accidents is increasing four percent a year.
We are also confronted with a very aging pipeline system. 24 percent of gas pipelines are now more than 50 years old. The section of pipeline inspected and involved in the Carlsbad tragedy was almost 50 years old and suffered substantial internal corrosion, and this Board found that it had never been properly inspected.
Congress and the Board have been aware of theunacceptable state of pipeline safety for many years, have made numerous recommendations, and have given the Office of Pipeline Safety at the Department of Transportation guidance and the steps needed to be taken.
Regrettably, under both Republican and Democratic Administrations, OPS has not been responsive until just recently. GAO found that the Office of Pipeline Safety had failed to implement 22 statutory directives for regulations and studies. 12 of these provisions go back to 1992 or earlier.
The Office of Pipeline Safety has had the lowest rate of any agency at DOT for compliance with NTSB recommendations.
In addition, the General Accounting Office has challenged the OPS policy of reduced reliance on enforcement fines. These new actions by OPS are encouraging, but there's a long way to go.
The Administration -- on November 3rd, the Office of Pipeline Safety issued an important pipeline safety rule, and the White House issued a presidential directive for more safety measures. They're encouraging, but those are only first steps.
If the past is prologue, progress will be made only if the public, the Congress, this Board stayon course with DOT and with the OPS and keep this issue publicly visible and insist that actions be taken.
The Final Rule issued by OPS deals with one segment of the industry, the larger liquid pipelines. The rule requires these operators to inspect and promptly repair pipelines in populated as well as environmentally-sensitive areas, and to take systematic steps to detect and repair leaks. Those are important.
Mandatory inspections will prevent future tragedies. The need for regular inspections is under-scored by the age of our pipeline system.
The company responsible for the Carlsbad pipeline tragedy never conducted an internal inspection of the pipeline involved in the explosion. Properly-conducted inspection would very likely have uncovered the corrosion problems before they led to a tragedy.
If you don't require pipeline inspections, there will be more tragedies. We must not have another Bellingham, another Carlsbad, another Edison, or another Moundsview.
In 1987, this Board recommended that OPS require periodic inspections. In 1992, Congress passed legislation that directed OPS to adopt regulations. It didn't by law enact them, directed the agency to adopt regulations requiring inspections by 1995.
On November 3rd of this year, 13 years after this Board's initial recommendation for periodic inspections, eight years after the statutory mandate, the Office of Pipeline Safety has finally issued a rule imposing pipeline inspection requirements for one segment of the industry.
Now, there are many desirable provisions. Requires periodic inspections, at least one every five years, that they be conducted by internal inspection tools or pressure tests. Inspection can be conducted by other methods, if the operator demonstrates that the alternative method produces an equivalent under-standing of the pipeline.
They have to notify OPS nine months in advance of making a change, and the rule establishes schedules for repairs of defects identified in the inspections, and the office has pledged in its rule to review all inspection plans and make changes where required.
That is backed up by the presidential directive, and it goes on to state that if inspection plans are found to be inadequate, OPS should use its existing legal authority to review -- to require revisions in the program, including requiring the use of internal inspection devices where appropriate.
That parallels legislation that I introduced in the House that was never acted on but doesn't go quite as far as we think should go.
While I generally support the OPS rule, one part of the rule raises important questions, which I would hope the Board will study carefully, particularly when that issue will have to be resolved in the rules affecting the "rest of this industry", and that issue is the deadline for the first required inspection.
The first required inspection has a deadline of only seven years, with the proviso that at least half of an operator's lines, representing the highest risk, must be inspected in three and a half years.
I don't see why we need that lengthy period of time when there can be failures, given the age of the nation's pipelines.
In the rulemaking, a number of comments submitted to OPS suggest a baseline assessment of five years or less. The Environmental Protection Agency, the Department of Justice, several cities, the NTSB, several environmental groups, objected to the seven-year time line for baseline inspections.
The office concluded, I think a faulty conclusion, that a seven-year deadline would result in a better assessment than a five-year assessment, thatinternal inspection in vendors would not have the human and mechanical resources needed to conduct these inspections during the next five years while meeting the current needs of the industry.
Mr. Chairman, in my many years of work on aviation safety, that is the argument raised again and again for not proceeding with ground proximity warning device, with traffic collision avoidance systems. We have heard it again and again. The industry can't gear up. They can't manufacture it fast enough. They can't do this quickly enough, and yet when forced to do it, when the market is there and necessarily so by law or by rule, the industry has responded, and lives are saved.
The OPS conclusion that there would be inadequate capacity for internal inspections over the next five years is based on a consultant's memorandum. The memo suggests even on its face that this conclusion is not definitive. The consultant says, "Getting a good handle on these numbers has been like pulling hens' teeth. These pigging guys are extremely protective of their data and sometimes misrepresent purposely so as to confuse their competitors." The consultant admits that his estimate of utilization rates is a guesstimate.
The key question of how much the industry would expand over five years, consultant made estimates of growth but gave no indication of the basis for his estimates.
When the Government adopts regulations requiring increased inspections of an industry, the industry will develop the capacity to conduct those inspections. The burden of proof should be on those who claim that the industry will not be able to expand to conduct the baseline assessments or produce the necessary equipment to conduct those inspections.
I don't think that the memorandum on which OPS relied satisfies this burden, and I think it should be revisited.
Fortunately, the issue has been taken to the level of the President, who has responded and pressed the Department of Transportation to act. However, I am concerned by the lack of specificity in the directives.
Of great importance is when OPS will issue regulations on inspections for the operators not covered by the November 3rd rule. That November 3rd rule covered the large liquid pipelines but not smaller liquid nor gas pipelines. For these operators, OPS hasn't even issued a notice of proposed rulemaking.
The President only required OPS to develop aplan by January 15th for adoption of rules for small liquid pipelines. A plan. Mr. Chairman, I think that the office should be required to issue an NPRM by January 15th so that we have something on paper, in place, for the public to respond to. A plan is insufficient.
With only a plan in place, there's every possibility that the new Administration of either party will want to step back and reassess the issues, and then this thing will drag out for another three or four years.
I'm also concerned about the difficulty OPS has had over the years adopting rules, and parenthetically, from a safety standpoint, Mr. Chairman, this is not just a problem with OPS, it is Department of Transportation-wide.
I asked the Inspector General of DOT to evaluate the performance of the Department of Transportation on rulemaking. The study confirmed my very worst fears, that we've gone backward, not forward. The IG found serious deficiencies in the rulemaking process, that the department is taking on average twice as long to issue rules as it did six years ago.
In 1993, DOT issued 45 rules and took anaverage of 1.8 years to complete work on each one. In 1999, the department issued 20 rules after taking 3.8 years on each one. It's taking twice as long to do half as much. That's a sad commentary, but it's like the Russian economy was. Not as good as last year but better than next year. They need to do better, and this Board helps by keeping the spotlight on their safety rulemaking inadequacies.
The Inspector General concluded, and I agree, that the problem is basically one of management. The existing process requires the concurrence of so many offices in DOT before rulemaking can go forward, that inevitably it gets bogged down.
The problem in rulemaking at DOT is that if any office disagrees with the rule, that office has the power to stop the process dead in its tracks. DOT's top management doesn't get adequate information about delays in rulemaking and fails to communicate to the responsible offices.
So, when we're talking about issuing a plan, about issuing future rules, all these hoops and hurdles that you have to go through to get a safety rule out, Mr. Chairman, are discouraging and disappointing, and in that scenario, I think this Board is the last best hope for forceful action, keeping the spotlight on thecurrent steps forward that the Office of Pipeline Safety has made, addressing the issues of technology, and assuring that there is clarity on the issues before us.
There is a role of prime importance here for the Board. It should be the counterforce demanding prompt passage of effective regulations, both within the department and within the Congress.
There's momentum now from the issuance of this rule on large liquid pipelines, but we need to assure that all of those steps are taken vigorously, and we need to work with all interested groups, and I'm certainly willing to do that, to ensure that we continue to make progress.
"A la mort, il faut la vrai". To the dead we owe the truth. The truth is we can make the nation's pipelines safer, faster, more effectively than we've ever done in history. This Board is on the right track.
Thank you, Mr. Chairman.
(Applause)
CHAIRMAN HALL: Thank you, Congressman.
Because RSPA Administrator Kelley Coyner has a scheduling conflict, we've offered her the opportunity to speak before we begin our staff'spresentation on the issues.
As I mentioned earlier, RSPA's Office of Pipeline Safety has begun to take action that will strengthen the nation's pipeline safety requirements, and I want to urge the industry to support RSPA's efforts to establish national pipeline inspection and testing standards.
Administrator Coyner, I appreciate your willingness to be with us this morning. I know that I appreciate, also, the fact that both of us traveled together to Bellingham and to Carlsbad, saw together and witnessed together the devastation of those two events, and I appreciate your commitment and your willingness to be here this morning, and we look forward -- the Board looks forward to hearing from you.
ADMINISTRATOR COYNER: Technology. What can I say?
I appreciate your creating this forum to examine both the state of technology for assessing and managing pipeline integrity and identifying those areas where we need to go forward.
Mr. Chairman, you mentioned our visits to Carlsbad and also out to Bellingham, which really underscored why this is so important.
The timing of this hearing is particularly helpful on the heels of congressional action in the appropriation area and the debate concerning pipeline safety reauthorization as well as President Clinton's announcement of significant actions to strengthen pipeline safety and environmental protection in this country.
As you know, I'm the Administrator of the Department of Transportation's Research and Special Programs Administration, which is responsible for the Office of Pipeline Safety. I welcome the opportunity to highlight the strategies we are taking to improve pipeline safety.
We have issued a Final Rule enacting tougher standards for pipeline integrity. This rule includes mandatory testing and strengthening our regulatory enforcement and research activities. The rule will improve pipeline integrity using existing technology and promote the development of enhanced technologies that we hope will provide even better tools in the future.
Our Final Rule strengthens protection for pipelines transporting hazardous liquids in populated and environmentally-sensitive areas. It is now complete. This integrity management rule is the firstof a series. It applies to operators of hazardous liquid lines that are 500 miles or more in length.
A second rule, which will follow shortly, will apply to operators of hazardous liquid lines which are less than 500 miles in length, and we expect to be very similar to the first rule.
A third rule will apply to the operators of gas transmission lines and will be published in the Spring. The rule requires operators to assess the baseline condition of their pipelines. The most high-risk segments must be tested within three and a half years, the balance in no less than seven years. We also require operators to perform periodic testing with an interval of not more than five years. Testing alone is not sufficient. This rule significantly raises the bar for safety by requiring operators to bring together information on all risks facing a pipeline system. Something we saw how critical it was when we were in Bellingham, Washington, last year.
We require operators to document those risks, lay out a plan to address those risks on a prioritized basis, and then implement the plan. Upon review of the plan, we will hold operators accountable. The risk for third party damage must be considered with the testresults to truly evaluate the significance not only of the baseline assessment but also the mitigation and prevention measures which must be taken as a result.
When operators have determined what additional preventative and mitigative actions are needed, they are required to assess the adequacy of their existing leak detection capability.
Leak detection capability may be the most important prevention action. The technology in this area is rapidly advancing, and the systems for assessment vary enormously.
In our review of operator integrity management plans, we will assess whether or not the operator has correctly determined the need to modify its leak detection capability to protect critical areas and enforce corrections that are needed.
We must make use of technology now available to maximize the protection of people and the environment from pipeline ruptures. At the same time, we must remain committed to improving the ability of tools to detect more types of defects with greater sensitivity and reliability.
Here are the areas we see immediate need for further research action. Transverse flux technology offers promise in detecting problems of seam welds. Itmust be fully field verified and documented.
Technologies to detect existing outside force damage and to monitor damage in real time as it occurs is a key priority.
Tools, such as sensing devices, to prevent damage from occurring in the first place that can be mounted on construction equipment to detect the presence of underground pipelines could be very beneficial.
Improvement in the capacity of leak detection technologies to be more sensitive to the size of the leak and to reflect the distinct qualities of individual pipeline systems. These leak detection systems must be less dependent on the human controller to increase their efficacy.
Consistent with the President's memorandum, we have been working with the Department of Energy for some time to coordinate national pipeline research by leveraging public and private resources to address pipeline safety and reliability issues.
We have been assisting the Department of Energy review responses to their recent solicitation proposals in the area of common interests, including third party damage, leak detection, enhanced inspection technologies, integrity evaluation and advancedmaterials.
By working together, we can maximize the results that enhance the integrity and the reliability of the nation's pipeline network.
Next year, the Research and Special Programs Administration plans to hold a national symposium, sponsored jointly with the Department of Energy, and I invite the Board to join us, to further highlight current research and development efforts, to clarify national pipeline research needs, and to identify the most effective and efficient means of meeting those needs.
I want to challenge the pipeline industry to invest resources in developing new and better inspection and detection tools and practices.
In closing, I would like to thank the Board for holding this hearing and help highlight the need for enhanced research and development efforts. The information generated by this hearing will contribute directly to the development of a national research program for pipeline safety and create a climate of innovation that will in turn lead to enhanced safety.
We know a lot about what will improve pipeline safety, but many unanswered questions remain. We must continue to be vigilant, whether there is anaccident or not, to ensure that we have the best, most effective tools in place to protect people from future pipeline ruptures.
I appreciate this time with you, and I'd be glad to answer any questions you may have.
CHAIRMAN HALL: Well, thank you very much, Ms. Coyner, and in the interest of time, the Board has assembled just a few questions for you. We know you're on a tight schedule.
Are many liquid pipeline companies now conducting internal inspections, and do you know whether they inspect all or just part of their systems?
ADMINISTRATOR COYNER: About 20 to 30 percent of the liquid pipeline companies conduct internal inspections, and I'd have to ask Stacey in terms of what the percentage of that means in terms of their entire systems. I think it's probably the bulk of their system.
CHAIRMAN HALL: Stacey, would you please identify yourself, since we have a court reporter? Your microphone's working fine. We just would like to have you identify yourself for the record.
MS. GERARD: I'm the Associate Administrator for Pipeline Safety.
I think our analysis shows that a great manyof them are testing, but it would have taken 12 years to complete the testing that this rule requires without the regulation.
So, we believe this regulation would approximately double the rate with which this amount of testing would be completed, and that would be the high-consequence areas.
CHAIRMAN HALL: Now, this is the liquid pipeline. How would that compare to the natural gas transmission lines?
ADMINISTRATOR COYNER: It's roughly the same in terms of the percentage of the lines.
CHAIRMAN HALL: Okay. Now, we have all referred to the fact that the President issued new regulations for pipeline safety which requires the companies to assess their pipelines, conduct inspections and tests, establish regular repair schedules and employ methods for leak detection.
Does RSPA have the resources necessary to ensure that the pipeline companies meet these new safety requirements?
ADMINISTRATOR COYNER: We do not. With 55 inspectors on hand, we are not currently in the position of reviewing these plans on a two-year basis to ensure compliance.
We will be working to fulfill the President's request that we identify what the necessary resources are and ask that that be included in the President's budget request next year.
CHAIRMAN HALL: And how is your agency funded?
ADMINISTRATOR COYNER: Our agency is funded through pipeline user fees that are assessed on the transmission companies in both the liquid and natural gas area, and we are also funded through a draw-down on the Oil Pollution Act Trust Fund as well.
CHAIRMAN HALL: Okay. What is the percentage of each? Do you have any idea?
ADMINISTRATOR COYNER: It changed somewhat this year. Let me defer to Stacey Gerard in terms of what the percentage is.
MS. GERARD: I think we're currently getting about $8 million of $47 million from the Oil Reliability Trust Fund.
CHAIRMAN HALL: Okay. But the agency is not funded similar to most of the other regulatory arms of the Department of Transportation through taxpayer dollars?
ADMINISTRATOR COYNER: It is not funded through direct appropriations, through the GeneralTreasury Account.
CHAIRMAN HALL: So, you are essentially dependent on the industry and this Oil Fund for your funding?
ADMINISTRATOR COYNER: That's right. Whatever the appropriation level is determines what the levels of the user fees are.
CHAIRMAN HALL: Okay. Well, some, as you know, have been critical that the new safety requirements may not be strong enough.
Can the pipeline operators meet the new requirements without doing internal inspections?
ADMINISTRATOR COYNER: The pipeline companies have to either do -- use a smart pig, which is what we typically think of as an internal inspection. They are to do hydrostatic testing or they must show an equivalent level of safety.
One of the things, I think, that's very important, and it's made very clear in the preamble, is that these plans must be reviewed and are subject to our directing them to use a better technology if we determine that's what's necessary in that area.
CHAIRMAN HALL: Okay. And when do you anticipate completing rulemaking to include all natural gas transmission pipelines and liquid lines less than500 miles, which are not covered in this rule, and I believe would not therefore have covered the Bellingham accident?
ADMINISTRATOR COYNER: It would have covered the Bellingham accident.
CHAIRMAN HALL: Would have.
ADMINISTRATOR COYNER: The 87 percent of all liquid transmission lines were covered by this particular rule.
When we devised this particular approach to the rulemaking about 18 months ago, we thought that the approach for small liquid lines might be substantially different. At this point, we believe it's going to be very similar, and we expect to have that rulemaking out in a matter of weeks.
CHAIRMAN HALL: What about on natural gas?
ADMINISTRATOR COYNER: On natural gas, we expect to have an NPRM out in the Spring of this year, and it will probably take the balance of the year to complete the rulemaking.
CHAIRMAN HALL: And the Final Rule addresses the need to test pipe in high-consequence areas. We both obviously were out in Carlsbad which would not fit under that rule.
How long a period of time would it take forthose pipes to be covered? That pipeline to be covered?
ADMINISTRATOR COYNER: We would expect that that pipe would be covered by including areas where people congregate. This is, as you might imagine, a very difficult area definition, but I believe that we are in agreement within the Government and also with the various stakeholders that we absolutely have to find a way to cover areas where people congregate.
This, as you know, Mr. Chairman, was an area that was an informal campsite but one that was very popular and well known, and we believe that it's important that we look to those places where people may congregate for worship or for recreational purposes or for any purpose and make sure that they're covered by high-consequence areas as well.
CHAIRMAN HALL: And, finally, does RSPA have any plans to support research to enhance the capability of pipeline inspection tools?
ADMINISTRATOR COYNER: We have an extremely limited budget for this at this point in time. It's a few hundred thousand dollars.
We hope to request substantially more in the coming year and to leverage our resources with the Department of Energy's research as well, but inactuality, there needs to be a substantial investment by the private sector, not only by the pipeline companies themselves but also by the technology sector, in order to achieve the goals that we've set for ourselves.
CHAIRMAN HALL: And I guess, in closing, Ms. Coyner, how many years have you been in this position?
ADMINISTRATOR COYNER: I was confirmed in August of 1998.
CHAIRMAN HALL: Well, I would like to personally thank you for your public service and for the leadership you've brought to bringing about action on many of the recommendations of the Board that have been longstanding, and I've appreciated your dedication in attempting to move those recommendations through rulemaking into regulation.
ADMINISTRATOR COYNER: Thank you, Mr. Chairman. This actually, I think, will be my last public hearing to participate in as Administrator of the Research and Special Programs Administration, and I believe that it is really an opportunity to lay down a marker about what else we need to do, not only in responding to your recommendations and congressional mandates but really maximizing the protection that we afford the communities along the pipelines and theenvironment, so that none of us are in the position of visiting communities, such as Carlsbad and Bellingham, in the wake of such a terrible, terrible tragedy.
I appreciate your focus on this and the rest of the Board Members' focus on this. Pipelines, a lot of times people say now, you're at the Department of Transportation, and is that a mode of transportation, and you all certainly understand that it is an important mode of transportation to meet our nation's energy needs but also understand that we have to do that in a way that is safe and protects the environment.
Thank you very much for your service as well.
CHAIRMAN HALL: Thank you very much, and Congressman Oberstar, and Ms. Coyner, thank you very much for joining us. You all are excused, and we appreciate your attendance and participation.
ADMINISTRATOR COYNER: Thank you.
(Applause)
CHAIRMAN HALL: We will now proceed with our public hearing, and I'll call on Mr. Bob Chipkevich for an introduction of staff and the program.
MR. CHIPKEVICH: Thank you, Mr. Chairman and Board Members.
On my left is Joe Kris, who's the Hearing Officer for this hearing. On my immediate right is Cliff Zimmerman, Pipeline Accident Investigator and IIC on some of the accidents we'll be talking about today.
Rod Dyck, next to him, who's the Associate Director for the Pipeline Division. Then, next to him, Jim Wildey, who is Chief of the Materials Laboratory in the Office of Research Engineering, who helps us extensively in examining pipelines that have been involved in accidents.
If the Chairman pleases, Mr. Dyck does have a presentation for the Board, when you're ready for it.
CHAIRMAN HALL: All right. Please proceed, Mr. Dyck.
MR. DYCK: Thank you, Mr. Chairman and Members.
According to the Research and Special Programs Administration or RSPA, it regulates over two million miles of natural gas pipelines and about a 157,000 miles of hazardous liquid pipelines.
This slide shows major pipelines in North America. It doesn't include natural gas distribution pipelines. The operation of pipelines with integrity problems has been a reoccurring issue in accidents investigated by the National Transportation SafetyBoard.
In 1987, as a result of investigations into three pipeline accidents, the Safety Board recommended that RSPA require pipeline operators to periodically inspect their pipelines to identify corrosion, mechanical damage, and other time-dependent defects that may affect their safe operation.
As noted, RSPA has completed a Final Rule on Integrity Management and plans to publish it in the Federal Register this month.
Accidents investigated by the Safety Board involving the operation of pipelines with time-dependent defects have continued to occur. For example, in 1994, in Edison Township, New Jersey, a natural gas transmission pipeline ruptured. The gas ignited, sending flames 400 to 500 feet upward and destroyed eight buildings.
Examination of the ruptured pipe revealed previous mechanical damage to the exterior of the pipe that reduced its wall thickness. A crag grew to critical size when it then ruptured. Contributing to the rupture were brittle properties of the pipe material.
In 1996, almost 500,000 gallons of gasoline were released into marshland near Grammercy, Louisiana,when a previously-damaged section of pipeline ruptured. This slide shows mechanical damage found on the pipe.
I can't get my pointer to work, but the Numbers 1 and 6 show scrapes, and the circled areas with dotted lines show dents on the pipe that were found, and these are -- this is the mechanical damage that we found. The contractor damaged the pipe about six months before the rupture occurred.
In 1996, nearly a million gallons of fuel oil were released into the Reedy River near Fork Shoals, South Carolina, when a section of corroded pipe ruptured.
In 1987, an in-line inspection device was run through this segment of pipe. The inspection contractor noted an anomaly at the eventual rupture location. The anomaly was assessed as a dent and was judged to require no corrective action.
In March of 1996, another in-line inspection device generated data that indicated pipe wall thinning. The Safety Board found that the subsequent efforts by the pipeline operator to measure the extent of the wall thinning were insufficient and did not reveal the full extent of corrosion damage. Before the corroded segment of pipe was replaced, the pipe ruptured during a pressure surge.
Also in 1996, a rupture of a pipeline near Lively, Texas, sent a butane vapor cloud into a residential area. The vapor ignited as two residents in a pick-up truck drove into the vapor cloud, killing both.
In May of 1995, an in-line inspection tool was run through the pipe, generating data that led to the conclusion that the rupture area only had light corrosion damage. The Safety Board found that rapid corrosion had occurred on the pipeline since the 1995 in-line inspection.
In 1998, a rupture in a pipeline in a landfill in Sandy Springs, Georgia, resulted in the release of more than 30,000 gallons of gasoline. When the pipe was excavated, it was found to be buckled and cracked. The Safety Board found that the pipeline ruptured because of settlement of soil and trash underneath the pipeline.
The Safety Board is currently investigating six other pipeline accidents that occurred during 1999 and 2000 that may also involve pipeline integrity problems.
In 1999, a pipeline rupture in Knoxville, Tennessee, released over 50,000 gallons of diesel fuel into the Tennessee River. A brittle-like crack wasfound on the pipe. The Safety Board is investigating whether corrosion initiated the crack and if the material's toughness had a role in this rupture.
Two days before the rupture, an in-line inspection device was run through the pipe segment with no anomalies in the rupture area reported.
In June 1999, in Bellingham, Washington, a pipeline accident released approximately one-quarter million gallons of gasoline, and three persons lost their lives. We found several areas of external mechanical damage in the vicinity of the rupture.
The arrow points to the gouge in which the rupture initiated.
In 1996 and '97, the pipeline operator conducted in-line inspection of the pipeline which indicated the presence of anomalies in the area of the subsequent rupture. The pipeline was not excavated in this area before the accident.
In January 2000, in Winchester, Kentucky, a pipeline accident released about 490,000 gallons of crude oil. Safety Board investigators found a dent on the bottom of the pipe in the rupture area.
In March 2000, in Greenville, Texas, a pipeline accident released about 565,000 gallons of gasoline. We found indications of cracking thatinitiated at the edge of a longitudinal seam weld.
In April 2000, near the Chalk Point Electric Power Generating Station in Maryland, a pipeline accident released about a 125,000 gallons of fuel oil. We found a crack and a buckle at a bend.
In 1997, an in-line inspection of the pipeline was conducted. The inspection report indicated the presence of a welded pipeline fitting at the approximate location of the bend. However, there was no fitting at this location.
In the case of an August 2000 natural gas pipeline explosion and fire near Carlsbad, New Mexico, that killed 12 people, we found significant internal corrosion at the rupture location. The pipeline segment that ruptured was constructed in 1950.
Today, we will focus on technologies available to assess the integrity of pipelines, such as the use of in-line inspection tools. We need to identify the benefits and limitations of these tools and to determine the status of on-going research.
Tomorrow, this Pipeline Safety Hearing will provide a forum to address the capabilities of pipeline-operating systems to identify leaks and provide sufficient alarms so that controllers can take timely action to reduce the consequences of leaks.
The lack of timely recognition that release has occurred has also been a reoccurring issue in accidents. For example, in the May 1996 accident near Grammercy, Louisiana, almost immediately after the rupture, several alarms sounded in the pipeline operator's control room, some showing that pumps had automatically shut down.
The pipeline controller said that he initially believed that the alarms resulted from refinery activities that had in the past generated alarms and which also automatically shut down pumps.
One alarm reported a line balance alarm, showing that the amount needed from one part of the pipeline differed significantly from another part. The controller said that he had anticipated a positive value from the line balance alarm because of the shut down of the pumps.
He said that he therefore did not read the full alarm message and did not note that the line balance alarm actually showed a negative value. The controller worked to restart pumps that had shut down automatically.
About an hour after the rupture, the controller received another line balance alarm. This time, the controller closely examined data and thenconcluded that a leak had occurred. Ultimately, 500,000 gallons of gasoline were released.
In the case of the November 1996 pipeline accident near Murfreesboro, Tennessee, a pipeline rupture resulted in the release of about 85,000 gallons of diesel fuel.
During the accident, a controller did not notice an over-pressure condition building against a closed valve at a pump station because the control room displayed an incorrect location for a pressure transmitter. The system recorded a sudden pressure drop at another pump station but no alarms occurred.
Although company procedures required shut down of the line in the event it was blocked, the controller continued to operate the pipeline and eventually succeeded in reopening the closed valve. He continued to pump diesel fuel through the ruptured pipeline for approximately one hour, until he realized that the expected pressure rise on the pipeline was not occurring.
The Safety Board is currently investigating five other accidents that may involve a delay in recognition of a leak. For example, in the February 1999 pipeline accident in Knoxville, Tennessee, the pipeline was not operating when the pipeline ruptured.
Records from the pipeline operator's control room indicated a sudden but small pressure drop at a pump station. No alarms were relayed to controllers. The pipeline was started up twice before controllers concluded that the pipeline ruptured, about four and a half hours after the rupture.
In the June 1999 Bellingham, Washington, pipeline accident, the pipeline operator reported that the computer systems became unresponsive because of inadequate computing capacity during the time frame that the rupture occurred.
Controllers did not recognize that the pipeline had ruptured and restarted the pipeline. About an hour after the rupture, controllers shut down the pipeline.
In the January 2000 Winchester, Kentucky, pipeline accident, the controllers shut down the pipeline about two hours after the rupture.
In the March 2000 Greenville, Texas, pipeline accident, at the time of the rupture, a pump automatically shut down. The controller didn't recognize the reason for the shutdown and started another pump in an attempt to keeping the line running.
In the April 2000 pipeline accident near the Chalk Point Electric Power Generating Station inMaryland, the pipeline operator for over one hour after the first indication of abnormal operation of metering instruments for monitoring the pipeline were not functioning during a maintenance operation that was on-going at the time.
Mr. Chairman and Members, the Safety Board staff has been deluged with requests to participate on panels. With just one day devoted to each topic, we simply do not have time to include all requests and have had to make difficult choices in choosing our panelists.
Therefore, we invited those that we could not include on panels to provide us with additional information for subsequent review and consideration.
For the panelists that are participating, we thank you for sharing your knowledge and experiences with us.
Thank you.
CHAIRMAN HALL: Thank you.
We're now going to -- before we move to the panels, we will take a short break, but if you would hold in your seats a moment, we will have four panels today.
The first will be on In-Line Inspection Service and will include representatives from Tuboscopeand PII. Then we will have a panel on Integrity Assessment, Pipeline Operators and Researchers.
I suggest we take a 15-minute break, return at 11:00, and we will then begin our panels, and we'll stand in recess until then.
(Recess)
CHAIRMAN HALL: We will reconvene this Public Hearing on Pipeline Safety, being conducted by the National Transportation Safety Board.
For our information of those in attendance, at the Board's website, www.ntsb.gov, we will have later in the day a webcast of the proceedings that you are now participating in, and we anticipate also having a transcript of these hearings available for -- on our website later as well.
So, if you have not had an opportunity to view our website, we certainly would encourage you to do so, and if you have any comments on how we might improve our website, particularly in the areas of pipeline and hazardous materials information, we would also welcome your comments.
We now begin the first of four panels that we will have today. Let me observe that, as Mr. Dyck commented, we have had a number of people that have wanted to present. We have had to be -- unfortunately,we have not been able to accommodate everyone that wanted to present, but we do have a number of presenters, and because we want to get through these presentations, we have a time limitation that will be enforced on the presentations.
So, Mr. Chipkevich, I will turn the hearing back to you. We will have this first panel of presentations, questioning by the staff and then questions by the Board.
MR. CHIPKEVICH: Thank you, Mr. Chairman and Members of the Board.
The next panel is the In-Line Inspection Service Panel. On this panel is Mr. Ravi Krishnamurthy and Mr. Tom Sawyer from PII North America, Incorporated, and Mr. John Parsons from Tuboscope.
Panelists, I would like to remind you that you have 15 minutes to give your presentation. The yellow light in front of you will go on when you have two minutes remaining. The red light indicates that your time is up.
Mr. Chairman, staff is ready to hear from the panelists.
CHAIRMAN HALL: Well, please proceed. Who will be going first? Mr. Krishnamurthy, please, sir.
Panel: In-Line Inspection Service
MR. KRISHNAMURTHY: Yes, I'll start.
I wanted to get my presentation on there. We're going to talk about pipeline integrity management, and we're going to try to attempt to present the inspection tools in context of the pipeline integrity management, and there is myself and Tom Sawyer, and I'll make the presentations, and we'll answer some questions.
I'm sorry, yes. Again, whenever you do a pipeline integrity management plan, you look -- you start with assessing the risk as the first step and of a system or a segment or sections of a pipeline. Then you attempt to identify potential failure modes. Is it corrosion? Is it cracking? And then, you look at in-line inspection tools or other mitigation methodologies.
Now, a critical factor in that is you want to match the appropriate inspection tool or technology to the potential failure mode, whether it's metal loss or weld deterioration, cracking, or dents, and a very critical aspect, next aspect, is post-inspection assessment.
You want to do remaining strength assessment or corrosion management or time-dependent modeling and longer-term modeling, but the in-line inspection has tobe put always in context of overall pipeline integrity management, and, of course, an on-going reassessment.
So, with this infrastructure in mind, I'll go into a little bit more depth on in-line inspection tools.
Now, really in-line inspection tools travel as products. They've been introduced since 1960s. They use very traditional technologies. Primarily, you can break it down into two. It's magnetic and ultrasonic. Pretty much all tools can be broken down into two.
They really record and store data. That's what they do, and they basically provide a snapshot of the condition of the pipeline.
Now, a key element to in-line inspection tools is quality of inspection, and that's where the difficulty in developing and analyzing the data, doing everything with these tools, come in.
Number 1 is detection. Next is discrimination. You can detect pretty much every anomaly in the line, but discrimination is very critical. Then you want to size these anomalies, sizing accuracy, and, last and the most important one, is repeatability.
So, in an in-line inspection designmanagement, you're looking at those four elements, all of them together, and when they come together, you will have a very good inspection and database at that point.
There are a variety of tools. You can look at geometry, bend, mappings, strength, but in my presentation, I'm going to primarily focus on defect detection and discrimination. That's all I'll focus on, and I'll focus on a few tools around that, and the different types of defects. I'll try to attempt to characterize our metal loss, cracks, laminations, etc., etc.
Now, there are -- remember, when we look at tools, you have to remember constraints. There are a lot of constraints with inspection tools. There are operational constraints in terms of what speed you can run this tool at, and how quickly, how much segment you can collect the data in, and there are a lot of operational constraints. Pipeline dimensions are a constraint. The bends are a constraint, and especially ultrasonic cleanliness is a constraint.
So, keeping this in mind, I'll go through every tool, identify what it can do, and what are the limitations, and what are some of the -- one of the most fundamental technologies people have used for ages is magnetic flux leakage.
It's the oldest of technologies being used. There are fundamentally two types. There is axial MFL, and then there is circumferential MFL, and circumferential MFL is what is referred to as a transverse of flux leakage, and it's basically circumferential, one is axial.
Now, really in an MFL tool, what you're looking at is a magnetic flux field imposed parallel to the pipe wall, and what you're looking for is the magnetic flux lines will be deflected if there's a crack -- if there is corrosion. Sorry. There's a metal loss, but if there's a ferrous material near the pipe wall or the properties of the steel change, you'll see deflection, and that deflection or that leakage is recorded by the sensor, which is that yellow spot there.
Now, at the bottom, there's a photograph of a magnetic flux leakage tool, a typical tool. You'll see the brushes represent the magnets, the magnets and the senses in the center.
Now, the length of these tools depending on the dimension you're looking at can range anywhere from 2.3 to 5.6 meters, and the velocity anywhere from half a meter per second to 5 meters per second. So, it's a range. 5.6, approximately 18 feet.
Now, (2) Specifications, and this is a very critical aspect of it, and these specifications normally are dependent on the diameter of the pipe.
As you go smaller in diameter, the specifications will deteriorate a little bit. Now, the minimal depth you normally can detect with this tool is about 10 percent wall, but I want to focus in on one aspect which I'll come to when I go to ultrasonic, is the pitting surface area, and what the axial MFL can do is actually detect a very, very small area, .28 by .28 inches, and are .4 times the wall thickness, and the sizing accuracy is 10 percent wall thickness, and again you can vary that. But I want us to remember that.
The other thing this tool does, it's been used in industry for a lot of times, and it effectively discriminates metal loss, splitting, circumferential defects, wall defects, but there's something this tool is not very good at. It is very poor at discriminating long longitudinal defects.
So, if you have a corrosion, which is long and stretches along the length of the pipe, this tool will detect the ends, maybe a few spots along the way, but it won't map the entire corrosion because it's an indirect measurement of corrosion. That's something we have to remember when we use this tool.
Another thing it cannot do is it cannot detect discriminate mid-wall defects. Now, these are some key limitations of this tool, but it's very effective at metal loss, pitting, circumferential defects.
Now, axial MFL tool comes with varying resolutions. You have, you know, low resolution, standard resolution. You have high resolutions. Then you have extra-high resolution, and what I've shown in the specifications here is what we would call high resolution.
Now, there's what we call extra-high resolution, also, but again that comes down to what value it adds at the end of the day, and again this tool, as far as our company itself, it's applied in over a 150,000 miles, and the industry has quite a bit of experience with this particular tool.
The next MFL technology tool we want to focus in on is the circumferential MFL. It induces circumferential magnetic field. Again, it's as opposed to the axial MFL. It goes around the circumference, and I want to show this picture here.
These two brushes represent the magnets and the sensor in the center. So, what you will do is you will have another segment, another section, like thisdownstream which will straddle the magnets. So, you will be at 100 percent circumferential coverage. So, this is basically a circumferential MFL tool. Again, length approximately 18 feet.
Now, this tool, it characterizes corrosion. It gets accuracy very similar to the axial MFL. In some cases, a little less, but it was primarily developed to detect seam weld defects, and what it does is it doesn't actually get you an exact depth, but what it allows you to do is to discriminate between crack-like and non-crack-like seam weld features.
But there's a key limitation on this tool which is very critical to note. It will only detect seam weld features or cracks which have an opening greater than .1 millimeter. This is very fundamental to this tool. It is because magnetic field is not disturbed, and we don't see a signal at that point for cracks.
It also detects and discriminates dents with and without corrosion and cracking, but exactly opposite to the axial MFL tool, the circumferential MFL is poor at detecting narrow circumferential indication. So, if you're looking for circumferential indications, this is not the right tool. Axial MFL is the right tool, and it cannot -- like the axial MFL, it cannotdetect or discriminate mid-wall indications.
Now, this is a much newer tool, introduced in the last two to three years. It only has about 3500 miles' experience.
Now, I just have an example of data from axial and circumferential there. If you look at the axial, you can see these spots, and if I read the axial MFL data in isolation, I would look at it and say these are pits. These are isolated pitting and maybe shallow. It may not impact the integrity of the line, but I go to the right side and look at the circumferential MFL data.
It actually characterize a very long-running corrosion, a long narrow axial corrosion, and that's where understanding what problem I'm looking at, what failure mode I'm going to look at is very fundamental to what tool I would use.
Ultrasonic -- again, ultrasonic is a wall thickness measurement tool. Unlike MFL, it's a direct measurement tool, and it's a compression wave tool which sends out compression ultrasonic wave and measures directly the wall loss.
Now, again, it also has two lengths of approximately 18 feet, and it comes in all different sizes, and this is a much more precise corrosionmapping tool. It actually maps the corrosion very accurately, whether it be isolated, pitting or long axial pitting.
But there is a down side to this, and I had mentioned in the axial metal loss tool, if you look at the size of metal loss it can detect, it can only detect and size metal loss greater than 20 millimeters. So, it is a limitation of the two when you run it at one meter per second and a certain pulse repetition frequency.
There are ways to improve that accuracy, but with the standard tool, you have this limitation. So, there is a point at which ultrasonic is very appropriate, and then there is a point at which axial MFL is appropriate.
But it is extremely precise at long axial corrosion, and, more importantly, it will characterize laminations. So, again I wanted to point out where these complement with the other tools.
Now, there's a very elegant tool for cracks, the ultrasonic shear wave. Now, this is a shear wave which is incident on the pipe wall at about 45 degrees, and like the compression wave, it comes perpendicular to the wall.
Now, this is a very, very, very sensitivetool, very accurate. It has 500 to 800 to 1000 sensors, depending on the size of the tool you're looking at.
Now, at this point, I wanted to point out that with all these tools, data interpretation is an extremely labor-intensive and a very difficult process. So, there's always a time delay in running the tool to getting the data analyzed.
In this particular tool, you generate 100 terabytes of data when you run this tool, and you go through an automated pattern recognition software to drop it down to 20 gigabytes, and from 20 gigabytes down on to usable data form, you're looking at manual interpretation.
But the value of that manual interpretation is it's extremely accurate in characterizing internal cracks, external cracks, and again I'm focused on longitudinal cracking. This can be reset for circumferential, but it's primarily used for longitudinal cracking, which is a problem you focus in on.
Again, there's a minimum defect line. There is a length accuracy. But again I want to focus in, it's very appropriate for stress corrosion cracking. It's very appropriate for fatigue cracks, weld defects,but again there's a size limitation.
It cannot go below 16-20 inches in diameter, and the other down side of every ultrasonic tool is extremely clean lines. That's a limitation. Much less restrictive in the MFL tool. It's actually been involved since '95-96, and it has over 5000 miles of experience. It's a very accurate tool.
Now, this is an example of a C-scan output, which is basically looking at an amplitude signal versus relative angle, and it has actually mapped the SCC cracks at this bottom with the C scan.
It's very valuable in characterizing these defects, but what I want us to remember is it characterizes length very accurately, depth estimates.
Now, there's another version of ultrasonic shear wave, which is for gas lines. Now, all ultrasonic tools require a liquid medium, and there's a tool which is modified using wheel probe for gas lines, where you have contained glycol, which allows you to use it in gas lines.
Again, I'm kind of running through it a little bit just to catch up on time, and I wanted to come to this particular tool. We are in the process of working on a new tool which is electromagnetic acoustic transducers, and this is focused for gas pipelines.
What the intention is, it should provide the same accuracy and ultrasonic shear wave that is provided for liquid lines. It uses a magnet to generate an ultrasonic sound through the pipe wall.
So, it will do the same things as an ultrasonic shear wave tool would do. This is in development. It's in R&D right now. We expect to have it out in a couple of years.
Now, again, I just want to summarize very quickly. Inspection tool selection, you know, I wanted to summarize to say when you're looking at metal loss, you're normally looking at three options: axial MFL, circumferential MFL, and compression ultrasonic tool.
In crack or crack-like defects, you're looking at shear wave ultrasonics or circumferential MFL.
So, really, when you go through this decision-making, you have to be very careful about how you decide what you need to run, if you need to run anything.
Again, inspection tools are very valuable for integrity management, but timing and technology is very critical to note. So, in terms of when you run this tool in the integrity life cycle for pipeline and the technology you use are extremely critical, and you always have to know the limitations, and it has to bein context of an integrity management plan.
Sometimes tools may not be the right answer. A pressure test may be a more appropriate approach, and we have to put it in that context, and post-inspection assessment is absolutely fundamental to having an appropriate in-line integrity management program.
Appropriate excavation data, good corrosion engineering mechanics, reliability of engineering has to be applied to this, for this -- for any of these tools to add value to integrity management.
Again, this timing of running inspection tools, to me, those timings should be decided based on time-dependent phenomena and inspection tools. A fixed timing, you always run the risk of either not running the right technology or sometimes when you run a corrosion after five years, your corrosion rates may be so slow, you're not going to see any difference between the two runs.
So, we have to be very cautious about how we time it and how we put it in context of an integrity management program.
I'm done.
CHAIRMAN HALL: Well, thank you very much, Ravi. That was an excellent presentation. A lot of information in a very short period of time.
We will have both presentations before the questioning. So, I'll ask Mr. Parsons if he would please proceed, and again welcome you, sir. We appreciate your participation in this hearing.
MR. PARSONS: Thank you, Mr. Chairman.
First of all, I'd like to say that the complexity of the tools has increased very rapidly in the last couple of years, and --
CHAIRMAN HALL: Mr. Parsons, these microphones are excellent, but they have the disadvantage that you have to pull them fairly close. So, if you would.
MR. PARSONS: Sorry. So, I'll -- if you can start the presentation. Okay. Thank you.
I think Ravi did a good job of talking about some of the complexities and the breadth of the new software packages and the tool systems that have been released recently.
So, what I thought I'd do is start off by looking at the true path system, which is our GIS package, and what that does is allows this data to be put together in a single platform and allow our analysts and our customer analysts to look at the data in a more rapid and more meaningful way, and to take a quick look at the tool fleet that we have today, andthen maybe take a quick look at the data analysis platform and some of the research we're currently carrying on.
The GIS platform basically allows you to look at the in-line inspection data in a real-world environment, and as you can see there, there's a wetlands area and a river, navigable waterway, which is traversed by a pipeline, and down here in the bottom left-hand corner is the elevation of the pipeline, and, so, you can see it crossing the river, and over here is the database that's supporting this information, and up on the top left is all the various types of information you can pull down on to a smart map.
On that map, the trajectory of the pipeline is actually placed there by the in-line inspection tool. It's not drawn, and all the features on that pipeline, every girth weld, every bend, is put there by the internal inspection tool itself.
The GIS work space itself, as you can see here, is able to identify lots of different features. For instance, here we have a school, and you can see the distance between these types of features, HGAs and USAs, in respect to the pipeline.
One of the abilities of this system is to measure distances between the pipeline and any HGAs orother school zones or whatever may be around the pipeline, and we can zoom in and out and measure those distances, and you're going to see right now we're measuring the distance between a housing development and the pipeline.
The system does have the ability to store data at every location. It allows you to pull down video of people on the pipeline. Here's somebody boring a hole along the side of a right-of-way. It allows us to access any type of information that the pipeline owner might gather, and in a second, I think you'll see a few more data items come up.
You can place reports on to the system. Here, we're taking a survey of a valve on the pipeline, and the exact location of that will be placed into the map. Here, you can see us observing a small river and the crossing across the pipeline. Obviously we can relate this to any anomalies that may be in the system after the fact. Here's a temporary launcher that was put on to the system to launch the in-line inspection tool.
Okay. One of the features of the software package is it allows us to interact between the GIS platform and the data from the in-line inspection tool, the actual evaluation data.
At the bottom of this system, you can see the database itself. Above it, you can see some anomalies on the pipeline, including the girth weld itself on the system, and you can scan or move by looking at database and applying different severity rules, sort in that, and then jump into those positions on either the smart map or the actual view of the data itself.
The current applications that we have on the platform is to provide a visualization tool for high-consequence areas and USAs. We can integrate this data with direct access information, CPs, CIS data. Any information from excavations that have been done on the pipeline can be put on there in order that you can make assessments of which are the most important anomalies to go to first.
We can interface this thing to a corporate GIS package that a pipeline owner may have, and we can also output directly to the National Pipeline Mapping System.
Future developments we have in mind is to apply a special query engine which will allow you to query on the system with simple rules, like give me all the positions where I have significant corrosion in relation to a river crossing or a school zone or a Class 3 area.
The second option that we're looking at is to put some risk assessment or integrity management rules into the system, and the third one is to make the system available across the web, to have a thin client capability, and last is to apply the system to field management directly.
Looking at the tool fleet that we actually run today, this is a typical tool that we run today. This one was released on to the market last month, and we'll see if the mouse now works. The front end of the tool is the drive module and includes the power system for the tool.
As Ravi mentioned, the second module in the tool is a large magnet that saturates the steel, and then there are some sensors between the two magnets which measure any flux leakage due to metal lost.
So, that section of the tool has another ring of sensors that discriminate between ID and OD corrosion. There are 256 channels on this particular tool to measure that data, and the third unit has the inertial navigation system in it and a measuring system to measure the distance along the pipeline.
The inertial navigation unit is a unit that applies the input on to the smart map which I showed you earlier, and the third ring of sensors is adeformation sensor ring which allows us to look at pipeline dents, wrinkled bends, and other features, besides corrosion.
We have similar features on our standard large diameter tools, but this is a new tool which we will be releasing the middle of next year, and this has some incremental features which are new to the industry.
The speed control has been out for a number of years now, and it allows us to control the speed, to get better data accuracy in gas pipelines. The second module is a circumferential magnetizer, which, as Ravi mentioned earlier, allows us to look at longitudinal narrow defects and to some extent cracking, and perhaps if you add to that the axial magnetizer in the same tool, you can increase the accuracy of the data for traditional corrosion or metal loss measurement.
We're also adding multi-access sensors to increase our accuracy, and the deformation sensors I showed on the previous page are also on this tool, and the data processing for these tools is significantly higher than tools on the market today, and we can run up to 4000 channels of data, and one of the other new features on this tool will be that we can look at the data directly in the field because it will be processedas it passes down the pipeline.
So, typically, a customer has to wait a number of days before he gets a report back from this. So, it will speed up the processing of the data, so they can act quickly if they find any significant anomalies.
This is a typical look at the data from the analysis system, and the top graph or view is of a saturated data on some spiral weld pipe, and the lower information is a reduced field, and in the reduced field, you can see there's a lot richer data which is due to the pipe itself not being saturated by the magnetizer, and you can actually see some of the mill defects and some high spots in the tool. This is actually a sample from our data.
So, the tribute package provides a very user-friendly interface for our analysts and our customers' analysts to use. It's a very accurate representation of the pipeline with regard to the data we have acquired, which is corrosion and deformation information, to date, and provides grading tools for anomalies.
We not only use the standard rules, we also do cluster interaction with things like R string, which I think Ravi also mentioned, and we can produce eithercustom reports or standard set of reports for our customers.
Development plans for the future. If the tools are run multiple times in the pipelines, we can
-- we should be able to predict whether particular corrosion area in the pipeline is active or inactive, and what the growth rate of the corrosion may be.
From the tool I showed earlier, we have multi-access, multi-magnet and multi-field strength sensors which allow us to increase the accuracy of the tools in the future, and we'll be able to provide improved strainer analysis in the very near future.
A quick look at R&D that's carrying on associated with that tool. The tools now will look at the magnetic field in multiple angles, and here's a quick look at some of the components of a -- here's a small corrosion pit in the pipeline.
As you can see, it's smeared from the circular component north and south in this particular view, and if we look at the magnetization field in the circumferential direction, we can then see the front and back of an anomaly fairly easy but perhaps not the sides, and in a third access, we can see that anomaly pretty well. So, by adding -- looking at all three axis of data, we can provide a lot more accurateresults.
If we then restrict the number of sensors we use just to a single sensor in the same plane as the magnetizer, and then look at this particular pattern here, we produce some holes in a piece of steel, and it lets it respond to the sensors.
As you can see, you see this smearing effect with an axial magnetizer, with axial sensors, which makes it difficult to discern holes themselves. With a circumferential magnetizer, the exact opposite happens, and you get smearing in the other direction, and that's why we believe if you take the two magnetizers shown on the previous tool, you can now see that pattern very clearly, and if you add the multi-access sensors, then it allows us to discriminate those holes even more clearly.
Looking at the mechanical damage and reduced field, this is a gouge that was placed in a sample in a pipeline, and this is a typical response from a saturated field. This is the response from the reduced field, which is taking into account the material characteristics of the steel in the pipe, and then if we subtract the two, we can see more clearly that gouge and perhaps the higher load created around that defect.
We still are not able to grade or discern thetype of defect, but we can identify that it's there, and that's what we hope to do in research over the next one to two years.
Looking at straight deformation-type systems, here's a rock dent on the bottom of the pipeline, and you can see that represented there, looking down the pipe and across the pipe.
Here is the internal combined strain inside the pipeline and the external combined strain, and we can see this quite clearly now in the system. Unfortunately, this is not accounting for structural steel changes due to mechanical damage.
Thank you.
CHAIRMAN HALL: Okay. Well, thank you very much, and we'll now move to the Technical Panel for questioning.
Mr. Zimmerman?
MR. ZIMMERMAN: Thank you, Mr. Chairman.
CHAIRMAN HALL: Please pull the microphone close, Mr. Zimmerman.
MR. ZIMMERMAN: Yes, okay. My first question I'd like to ask to Mr. Krishnamurthy.
MR. KRISHNAMURTHY: Murthy.
MR. ZIMMERMAN: Murthy.
MR. KRISHNAMURTHY: Ravi's fine.
MR. ZIMMERMAN: I'd like to know what is the confidence level that you now have in your cracked tool data and the evaluation of that as far as being commercially viable.
MR. KRISHNAMURTHY: Again, which tool --you're referring to the ultrasonic shear wave or which one are you referring to?
MR. ZIMMERMAN: Well, we can talk about --
MR. KRISHNAMURTHY: Okay.
MR. ZIMMERMAN: -- any of the tools that you have, yes, at this point.
MR. KRISHNAMURTHY: Okay. Again, the ultrasonic shear wave -- let me go to that one. That's the easy one.
It's very highly reliable. It's been documented to have confidence intervals of 90-95 to 100 percent in terms of locating linear indications.
Now, all of that -- it's -- I have seen documentations of about 80 to 85 percent of identifying whether it's SCC or some other type of anomaly, but now if you go to the TFI or the circumferential MFL, we are still more -- that is where the demarcation comes in.
When you look at that greater than .1 millimeter defect in identifying a seam weld that's crack-like, we're looking at about a 75 percentconfidence when you see where the seam weld is crack-like.
Now, we can identify all seam welds, whether it's crack-like or non-crack-like. That discrimination is much harder. So, there, we're looking at a much smaller confidence interval.
So, it depends on what kind of defects you're looking at and that.
MR. ZIMMERMAN: Okay. And then, the next point now, first, you have to find them obviously, but then your evaluation of them to determine which defects are significant, so that a pipeline company will know which ones that they need to repair. That's the critical part that I see once you can find a defect.
MR. KRISHNAMURTHY: Yeah. Again, let's go back to the shear wave tool.
The shear wave tool will very easily identify which ones are -- which -- what I would call critical cracks, which will fail by a traditional fraction mechanics analysis or whatever.
Now, what they -- the subcritical cracks, again they'll meet that threshold, 30 millimeters, and greater than one millimeter in depth. You can be 90-95 percent confident that it'll meet that very accurately.
Now, coming back to the TFI, TFI willidentify all seam weld features because you'll see an indication in the traces or in the control field they look at.
Now, the question is when they discriminated, they're only 75 percent confident that it's crack-like. That is where I'm talking about.
MR. ZIMMERMAN: Okay. And out of those indications, what kind of a report do you then prepare for the pipeline company so that it can act on the ones that are problems?
MR. KRISHNAMURTHY: Okay. Again, we saw them -- again, going back to the circumferential MFL, we saw them by what we would call crack-like and what we would call -- they call them -- we are 75 percent confident that it's crack-like, and we also give another category which we say we're 50 percent confident that it's crack-like. We're not sure about this one, and then there are seam weld features. So, we kind of categorize them.
In the circumferential MFL, we don't quantify the depth. It just meets specification depth, whereas when you go to the ultrasonic shear wave, you get an actual length, estimated depth, categorized depth. So, it's a different kind of reporting structure than shear wave.
MR. ZIMMERMAN: Okay. So, here again, I believe we're -- I'm not sure you're answering my question.
If we have -- we know to whatever confidence level you can determine these cracks and the depths, what kind of a report do you then put out to the pipeline company that indicates which one -- let's say you have 10,000 indications.
MR. KRISHNAMURTHY: Hm-hmm.
MR. ZIMMERMAN: Maybe five of them are significant. Maybe 200 of them are significant, that they should go out there and look at them and examine them.
How do you make -- how does your analyst go in and make that evaluation?
MR. KRISHNAMURTHY: Okay. Now, the analysts don't make the evaluation that they should go out. What they'll identify is these already deep cracks. We think they're long.
Now, that -- in the report, they'll identify these as significant cracks or, let's say like you said, out of 10,000, five are significant, they will be identified and highlighted, and then the rest will be categorized or summarized based on whether they're crack-like or non-crack-like or where they fall inthat. So, they are identified in the report.
The same thing in the shear wave tool. If you see some really bad cracks, absolutely, you won't wait for the report. There will be a direct communication on that. But it will be identified in the report as significant cracks or cracks which we think are significant or the analyst thinks is significant.
MR. ZIMMERMAN: Okay. For Mr. Parsons. A similar question. What's the confidence level in using your crack tool for commercial applications and evaluating the inspection data?
MR. PARSONS: Currently, we only have a prototype tool. It's still in the research phase. We took over a project from another in-line inspection vendor that was exiting the business, and when we developed or took that tool out on some preliminary runs, we found that it did not grade as accurately as had been advertised, and we have taken it back to the research phase and are redeveloping the sensors, and we expect that tool to take about two years to develop.
MR. ZIMMERMAN: Okay. Thank you.
The Board has seen cases where anomalies have been misinterpreted in both MFL tools and ultrasonic tools. I'd like to address my first question to Mr.Krishnamurthy first, and can you tell us if the evaluation of signatures of anomalies can be improved, so that we don't miss the ones that are critical?
MR. KRISHNAMURTHY: Yeah. Let me go back to your previous question, which is defects missed or defects misinterpreted or miscalled or called wrong or not even called in some cases.
There could be three or four reasons for that, and I want to go back to my presentation. One reason is the appropriate tool for the appropriate application. That's very fundamental to this. No one tool is a panacea for all problems. It identifies the predominant problem and use the right inspection, and in the context of right integrity management plan. The tool is only a subset of that plan.
The other aspect to that is a lot of these inspection technologies, there's an element of this which is automated, and there's a huge element which is you have a bunch of very dedicated analysts, like 50-60 people, sitting in offices, looking at data day-in and day-out.
So, there is a human element to it, which is why you'll see sometimes the confidence interval is reflected by virtue of the technology in some cases, and in some cases by virtue of the fact that it's ahuman interaction. So, those are two or three possibilities.
Now, there are ways to improve it. The first way, I think, to improve it is to put -- always put a tool in context of integrity management plan, and understand what you're looking at, understand what kind of corrosion you're looking at, not just going and say I'm going to run a UT tool for every five years or every eight years or run an MFL tool for -- we have to have an understanding of why we are doing that. That is one way to improve it.
The other way, which companies, like PII, in our Inspection Division, are working on it is, of course, improved training a lot more people. I mean, by virtue of the fact that as the load increases, you know, we are not always positioned. So, we have to be proactive about lower management in that respect.
Thirdly, improvement of technology, and the EMAT is an excellent of that, where it will be more accurate than sizing cracks in gas lines, for example, than the traditional -- Tom, did you want to add?
MR. SAWYER: Yeah. I think, just to add on Ravi's, I think there is an opportunity for more rigorous testing using facilities in the industry.
In looking at a multitude of defects andsamples from industry, which we can accurately characterize the tools against and build up a defect library much, much faster than we currently we do right now.
As I say, to show new technologies, you have to build up the track record, and there is feedback required from the industry, and the better feedback we get, working with the industry, the faster we can improve tool accuracies and confidence intervals and probabilities of detection.
MR. ZIMMERMAN: Well, that's another question. So, before I go on to Mr. Parsons, and now I'll get to you on this one.
Tell me a little bit about what kind of feedback you're getting from industry. Is it sufficient? Could you characterize that for us?
MR. SAWYER: Well, it varies, of course. I think in some of the newer technologies, we've been quite pleased with the feedback with the Transcan tool or the circumferential tool.
Because of its newness and seeing many things for the first time with this technology in situ as opposed to a laboratory environment or what we call a "pool test", the feedback has been tremendous in most cases.
In areas where we can improve is with some instances where there is no feedback. So, we also need to perhaps work more jointly with industry in terms of setting up of test programs where we can evaluate situations with depth samples.
Again, as I say, it can improve, but for the most part, the new technologies, it's been excellent.
MR. ZIMMERMAN: Okay. And then, in general, in your work course tools, the majority of inspections that you do, is it possible for you to, you know, set up some communication with the company so that the differences that they find would be reported back to you, so that you can again use them to further define your evaluation of defects?
MR. SAWYER: Yes. We use that on a case-by-case basis. It is dangerous sometimes to take a very specific example and then adjust your algorithms and your mathematical models and apply it across all the board.
We do on a case-by-case where there are difficulties, and we do verification digs in many instances, if not most instances, and we use that feedback mainly to adjust for that specific run or that specific inspection, but we don't necessarily apply it across the board entirely because there are dangers indoing that.
We are embarking on some initiatives in the future to where we would like to see industry user groups in which we have -- bring together multiple users of a certain technology to discuss what they have found, so we can get a more broader picture or holistic picture of what the tools' capabilities and limitations are.
MR. ZIMMERMAN: Okay. Mr. Parsons, could we go back to this first question then? Can the evaluation of the signatures of anomalies be improved, and how do you go about doing that at your company?
MR. PARSONS: Yes, they certainly can, and there are two things that we can do.
One is to improve the algorithms we've developed from the software side, and to model those and to build samples and verify the models, and we are actively working on that now.
The second thing we can do is to apply multi-magnet technology to the systems. I think as Ravi said, the circumferential tool is very good at narrow axial cracks, and the traditional tools are better at typical pits and general corrosion, and if you put those two things together, then you are going to get a higher-accuracy tool, but it is a much more complextool, and it's a more expensive tool.
So, we can improve the accuracy but not without cost.
MR. ZIMMERMAN: As usual, yes. Thank you.
Rod, would you like to -- I'm going to pass the questioning to Rod Dyck now.
MR. DYCK: Yes. This is directed to the entire panel.
What constitutes a pipeline that would be capable of accommodating these in-line inspection devices versus pipelines that can't?
MR. SAWYER: That's a very general statement. It's difficult to answer it succinctly.
There are certain restrictions in terms of bends, is one of the single largest restrictions. These tools typically are only capable of negotiating through pipelines in which the bends are not tighter than one and a half diameters of that pipe. The bend radius is not less than one and a half diameters.
Things, such as reduced port valves, can obstruct certain technologies. Other technologies, there are tools available that do -- that are called multi-diameter pipelines.
I know Tuboscope has developed such tools. We have also developed tools which can inspect multi-diameter pipelines. There are other -- a host of other features that make it impossible to detect through certain parts of the pipeline as well.
Line cleanliness is always a factor, as Ravi mentioned, particularly with ultrasonics, where you need direct contact with the pipe surface.
MR. DYCK: The entire panel, that constitutes your answer.
Could you talk just a little bit about putting these devices through hazardous liquid pipelines versus gas lines? What the differences might be, and what the end result might be from this activity?
MR. PARSONS: On gas pipelines, we are more concerned about the compressibility of the product. So, if there are reduced port valves or other restrictions in the pipeline, the tool will tend to surge, and, so, that's why we've built speed control systems to allow gas to pipe through the tools, and that's really the major difference, I would say, between a liquid pipeline and a gas pipeline, although certain products are quite corrosive, and, so, we have to be careful about the type of compounds and seals we use to ensure they don't fail in the pipeline, in the liquid pipeline.
MR. KRISHNAMURTHY: And just one other point after John's point, is the liquid lines normally offers more flexibility in terms of tools because at least, for example, in a liquid line, you can go to ultrasonics, which is very easy to do. In gas lines, it's extremely difficult to do that. It's a huge operation in fact. So, that's a big impact.
MR. DYCK: Is there a difference in the end result between the liquids and gas?
MR. SAWYER: No. There would be none, other than, say, line cleanliness would be the issue, but magnetics, for example, works -- is not affected by the medium whatsoever.
MR. DYCK: Thank you.
CHAIRMAN HALL: Mr. Wildey?
MR. WILDEY: Yes. I just have one question area for both of the companies here.
I'd like for you to comment a little bit on the interaction between the in-line inspection companies, such as yours, and the operators. Is there a synergistic thing going on, where are they a part of the evaluation of the data, and do you provide feedback to them perhaps on the frequencies of inspections? What can you comment on that area, please?
MR. KRISHNAMURTHY: Yeah. Again, as partof -- there is a lot of synergy. Like Tom mentioned, we get a lot of feedback on different systems and different tools, especially the new ones, quite a bit.
But the part of it where PII is moving in the direction, where we are getting more into integrity management and assisting operators in doing tool selections and providing frequency predictions. Absolutely.
Like I said, it has to be a holistic approach, and absolutely that's something we do, and we are continuing to do that.
MR. PARSONS: Yeah. I think Ravi's hit the main points. I don't have anything to add.
MR. ZIMMERMAN: Okay. Thank you, Mr. Wildey and panel. Mr. Chipkevich?
MR. CHIPKEVICH: For anyone on the panel, just a couple of brief follow-ups.
On the issue of obstacles that would prevent a pig from being operated to the line, other than bends in the pipeline, I guess that would include a bend, either horizontally or vertical because of hills and things of that nature.
What other type of obstacles do you face? Because, you know, the Board's been pushing for internal inspections and the testing of pipelines forsome time, and we do get answers back that a lot of pipelines can't be inspected.
You know, what are some of the other obstacles, and, you know, how difficult is it to change some of these pipelines so they can accommodate the tools?
MR. PARSONS: I would say the biggest single issue would be plugged valves in gas pipelines. They're extremely expensive to replace, and clearly the pigs can't pass through them. So, they'd have to be removed, and in terms of multi-diameter capacity, we, as a company, don't have a lot of those tools. So, we would have to build a new fleet of tools to handle a lot of reduced port valving.
MR. CHIPKEVICH: But there are tools that could be --
MR. PARSONS: We do have tools that handle reduced port valves, yes.
MR. SAWYER: We also have tools that handle not necessarily specifically port vales but multi-diameters in which -- so, any restriction on pipeline inspection is sometimes the availability of traps either not built into the system, and with looping of new systems and expanding the pipeline infrastructure, we have worked with vendors to build tools that willinspect more than one diameter, so that they don't have to spend the money on adding infrastructure which again is costly.
MR. CHIPKEVICH: Would you all have an estimate on the amount of pipeline by percentage is piggable today?
MR. PARSONS: Not really. Don't have a handle on that.
MR. CHIPKEVICH: What about a comparison of pipeline runs? We've had some accidents we've looked at over the years where there was an anomaly in a pipeline at a particular location, and then a subsequent run of that line showed a change in that anomaly.
Is it difficult, when you've got a pipeline that's a thousand miles long that you're examining, to identify and to compare anomalies that you see in a pipeline so you can give that information to the operator?
MR. KRISHNAMURTHY: Again, there are a lot --it is -- it can be done, but it's very difficult. Occasionally, it can be difficult. For example, you look at technology and what you have collected.
In ultrasonics, it's a direct measurement. So, you can do it a lot easily. I mean, if there isvalue, you do it. Especially on an overall integrity management program, there's value because it can assist you in predicting when you need to run your next inspection.
Going back to MFL, the way the data is analyzed, you have a clustering definition in MFL, and because of the change, as time-dependent corrosion happens, there is a change in the clustering. So, next time the analyst looks at the data, he'll cluster the data a little bit differently.
So, it's very difficult to do a direct measurement looking at tool reports and saying which --has it grown or not? But what we do normally is look for specific cases. We use the raw magnetic data and go joint-by-joint and subtract the signals and look for corrosions, and that's one way we overcome that problem.
Again, we don't -- we try not to do that for every defect. We try to do it in areas where we suspect there is excess of corrosion or some sort of thing. So, it can be done, but that's --
MR. SAWYER: Yeah. That technology is available now and was introduced to the marketplace earlier this year. As Mr. Krishnamurthy said, we analyzed the signal match, raw-to-raw signal, whichremoves the operator interpretation uncertainty and goes back to the accuracies of what the tool is capable of. It's raw signal matching.
So, that software is now available and those algorithms are available for all the magnetic technology within PII.
MR. CHIPKEVICH: And just one final question, and that is, the difference in inspecting gas lines and liquid lines.
You mentioned some of the tools couldn't be used, such as the ultrasonic, in a gas line, but what about the other tools? Can you inspect them with the gas lines? Can you get as good of data, and can you do the test without filling the gas line with liquid?
MR. PARSONS: From an MFL perspective, there is no difference between the two. In fact, typically the gas pipelines are easier to get the data because we don't suffer from things like waxing that we might get in a crude oil line. So, technology works very well.
MR. CHIPKEVICH: Thank you very much. Did you have --
MR. SAWYER: Well, the -- on the crack detection technology, as Ravi mentioned, there is a shear wave, which is designed for liquid, and there's a shear wave that's designated in a fluid-filled wheel,and that typically doesn't have the same level of probability of detection or the same level of confidence interval and accuracy that there is for the gas.
There is crack detection technology available. The newer technology is the EMAT technology, as Ravi mentioned, which is due to come out in the year 2002.
MR. CHIPKEVICH: Okay. Thank you.
Thank you, Mr. Chairman.
CHAIRMAN HALL: Okay. We'll move for questions up to the Board Members. Member Carmody?
MS. CARMODY: Yes, thank you.
Good morning. You've had the Technical Panel. You're now having the non-technical questions.
My first one goes to Mr. Krishnamurthy. I was interested in your comments that an integrity management program was crucial to determine what it is you're looking for and then select the tool.
With that philosophy, I would assume that you could use a number of different tools then on one pipeline, depending on what you were looking for.
First, my question, is that correct, and, second, is it common?
MR. KRISHNAMURTHY: Yeah. Absolutely. Thatis correct, and it is common whenever you look at a pipeline, and I have seen a few cases and have been involved in cases where you have corrosion and cracking. If you have corrosion and cracking, you have to look at -- you want to look at your technologies.
But normally, you don't look at your technologies. If -- again, when you're doing an actual risk assessment, you can come out and say this is the most predominant mode of failure, and in terms of risk -- I want to first address that.
So, you focus in on that and address that and come back, reassess and see if my next trend may be a different tool, depending on what I'm looking for. So, it can be staggered that way, too. So, that may be a better way of managing it.
MS. CARMODY: Okay. Thank you.
The training. I was also struck by what you said about the difficulty of interpretation, the difficulty of, I guess, training.
MR. KRISHNAMURTHY: Yeah.
MS. CARMODY: How much training does it require for a typical analyst to be able to interpret data?
MR. KRISHNAMURTHY: It depends. For example, MFL, it could be two-three months, but when you look atshear wave crack tool, it's normally a year's worth of training before we can get them up to speed. So, it's not something you can go out and -- we cannot just hire five people and bring them in. There is intensive training, and there is an experience level.
So, then even though we have base analysts and every analysis is checked by a senior analyst. So, there's a process involved there in order to capture this human element which I talked about. So, there is quite a bit of training, and again it depends on the individual, but it's training, and it can get tough on the folks because it's a very -- it can get very boring, I would say, because they look at like thousand miles of pipeline, and they're looking at every spot. So, it's a very tough job. It's an extremely hard job.
MS. CARMODY: Well, you touched on my next question, which is, I assume you have multiple people looking at particular interpretations. So, you have kind of a consensus opinion on --
MR. KRISHNAMURTHY: Correct.
MS. CARMODY: -- something like --
MR. KRISHNAMURTHY: Yeah. And in something like a shear wave crack tool, it actually goes through three or four people before it gets cleared and because of the complexity of interpretation there.
MS. CARMODY: And that would be before you send it on to the --
MR. KRISHNAMURTHY: Yes, yes, yes. Absolutely.
MS. CARMODY: I went out to the Carlsbad accident with Member Hammerschmidt and with Mr. Zimmerman. So, I have a special interest, I guess, in that one.
I know El Paso had said they could not inspect that particular pipeline with an internal device because of, I think, fittings, pipe fittings on the -- that went over the bridge.
Is there a tool in development or a tool you know of that might be useful in a pipeline like that?
MR. KRISHNAMURTHY: Tom?
MR. SAWYER: I believe it was an issue that was -- it was a couple of issues regarding fittings. I'm not intimately familiar with the specifics. I mean, some technical problems are surmountable, some have physical limitations which you just can't get around. So, I don't know specifically about that particular pipeline, no.
MS. CARMODY: Anyone else want to speculate?
(No response)
MS. CARMODY: Okay. I think that's all forme. Thank you very much.
CHAIRMAN HALL: Member Black?
MR. BLACK: Just a couple of questions, I think primarily about data.
Do you have -- say on a 10 or 15-20-year old pipeline, do you have any idea how many hits, in other words, how many spots need inspection on the data return per mile or per foot or whatever?
MR. KRISHNAMURTHY: Again, see, that would depend on the -- even if you fix the age, it will depend on the type of coating, type of service, type of environment. It's a very difficult question to answer.
Some lines would be one per mile, some may be one every 50 miles. So, it cannot -- it depends on the operating conditions. It depends on -- that's kind of why I would look at that risk assessment or integrity management plan to understand why that pipeline needs inspection.
MR. BLACK: But we're not talking about hundreds of hits per mile. We're talking about --
MR. KRISHNAMURTHY: Again, it depends on the pipeline. You could have pipelines which could be --
MR. BLACK: There's disagreement there or --
MR. SAWYER: Well, no. We've -- in our experience, I think Tuboscope would confer that thereare pipelines that are very old, that are extremely clean in terms of the number of hits, and there are pipelines due to the factors Ravi talked about where we have seen literally millions of significant features.
MR. BLACK: I guess that goes to the next question. With the location, with this inertial navigation system, what is the location accuracy?
In other words, if you're going to have your client go out and dig down, can you get it to the nearest foot or the nearest meter or whatever you're --
MR. PARSONS: Yeah. That's about right. About the nearest meter is typical. It depends a lot on the products in the pipeline and the tortuosity of the pipeline itself, how many bends it's got in it, but typically, less than a meter for a dig, for an anomaly.
MR. BLACK: And t