Multiple structure fires sparked by suspected gas line failure in towns north of Boston 7

[bimr]

Authorities ordered residents to leave their homes immediately after dozens of house fires broke out in a string of communities north of Boston Thursday evening.

The Massachusetts State Police were evacuating multiple neighborhoods and restricting access to the area.

"Residents in the affected towns of Lawrence/North Andover/Andover who have gas service from Columbia Gas should evacuate their homes immediately if they have not already done so," state police tweeted.

"Gas lines are currently being depressurized by the company it will take some time."

Link

 

[zdas04]

I wouldn't be surprised by more of this sort of problem in the future. No pressure reducing valve can be assured to be bubble tight in service, when shut hard some portion of them will leak. Historically this hasn't been a huge problem because we had pilot lights that created a non-zero minimum usage. Pilot lights are now considered "ecologically unsound" and new ranges, water heaters, gas fire place inserts, and HVAC units are pilotless. That means that a tiny leak-through of a regulator has no place to go and can put 30 psig from the street into homes rated for less than one psig. Most systems will not fail at 30 psig, but some will have components under pressure that really can't take 30 psig.

The failure in Boston will be described in the accident report, and we'll see how close the descriptions above are, but I don't think that this kind of failure actually requires much overpressure on the feeder trunks to happen as homes rush to zero minimum usage.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[Compositepro]

I believe all gas service regulators for buildings are self-relieving, that is they will vent gas if the downstream side goes much above setpoint. I think here the problem was that the supply pressure went way above 30 psig.

 

[bimr]

The only problem with that scenario is that the gas pressure never went to 30 psig.

"The senators noted that a federal regulator, the Pipeline and Hazardous Materials Safety Administration (PHMSA), reported that the pressure in the Columbia Gas system “should have been around 0.5 pounds per square inch (PSI), but readings in the area reached at least 6 PSI.”"

Link

 

[hokie66]

More and more confusing. The statement linked by bimr says that the National Transportation Safety Board is leading the investigation. You would think they have enough on their hands with aircraft accidents and bridge collapses. Pipelines transport gas and liquids, but the NTSB just may be getting overloaded/overpressurized.

 

[davefitz]

I heard recently that the Boston area nat gas distribution system is very old ( over 100 yrs old) and has old cast iron componetns that needed to be replaced. As a result the entire neighborhood supply piping will be operating at very low pressures ( <1.0 psig) and there will not be a pressure regulator at the external flow meter . The only diaphram pressure regulators are those at the individual household appliances. Those appliances cannot withstand a 6 psig gas pressure without failing. This is based on a recent statement by a gas utility spokesman for a more modern west coast location, and I cannot confirm its validity.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[hacksaw]

Boston Globe:

"The construction work involved replacing aging pipes in the area with plastic lines. But the National Transportation Safety Board said the utility failed to tell the construction crew about disconnecting or relocating the sensor, allowing the device to detect a drop-off in pressure in the abandoned line and signal to a nearby control station to increase the flow of gas into the system."

Apparently the control system did not include safety checks, pre-commisioning checks.

 

[itsmoked]

Geesh! Obviously nobody-in-charge.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LittleInch]

Here's the official preliminary report.

https://www.ntsb.gov/investigations/AccidentReports/Reports/PLD18MR003-preliminary-report.pdf

Seems almost inconceivable that the design of the modification didn't realize that the sensor line was big located nor that the regulators didn't have independent secondary pressure protection devices for the downstream sections. Nor that the monitoring centre had no executive commands available to it to remotely close the regulator station.

I had some issues before with this line of thinking which is different from an industrial design where the default is fail safe or fail close.

Distribution systems which shut off cause more hazards and disruption, but still need to prevent over pressure.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[RVAmeche]

Certainly a black-eye for Columbia Gas considering it was their own design and construction plan.

 

[John2025]

Yikes, A system that fails that catastrophically on the loss of a sensor is inconceivable. Surely there have to be backups, sensors fail all the time.

 

[pellucidar]

Somerville (a city only slightly north of Boston) has postponed some sewer work in my neighborhood because "n response to the recent explosions in the Merrimack Valley, the Massachusetts Department of Public Utilities ordered all gas utilities to halt the tying in of new gas mains to the existing system".

I also saw some recent numbers on the repairs from channel 25.

 

[LittleInch]

An update courtesy of my alter ego BigInch:

There has been some further developments since. A criminal investigation that was opened as of Nov. Hard to believe that nobody looked at a PID before cutting in.

https://fox61.com/2018/11/01/feds-open-criminal-probe-into-massachusetts-natural-gas-explosions/

NTSB has also issued a Safety Recommendation in response to their initial investigations of the Lawrence disaster.
(

https://www.ntsb.gov/investigations/AccidentReports/Reports/PSR1802.pdf)

It advises that only registered professional engineers (PEs) be allowed to work on utility facilities.

There has long been exclusions for PE requirements to work on utility installations in many states. Massachusetts being one. Texas another.
This accident resulted in my opinion from extremely bad engineering practice, well almost none exactly, plus again in my opinion, a poor design.

It also prompted me to write to NTSB.

==========================================================
As a former Texas P.E, now retired, but having worked a lifetime in pipeline design engineering, construction and operations worldwide, I could not agree more with NTSB's recommendations as recently expressed in "Safety Recommendation Report: Natural Gas Distribution System Project Development and Review (Urgent)" [doc number PSR1802]. Specifically I could not agree more with the cautions that NTSB expressed therein, to use only qualified professionals and the recommendation that Massachusetts eliminate the exception requiring a competent registered professional engineer, authorized within the jurisdiction, to approve utility plans. Frankly, in my opinion, that accident was the result of complete incompetence. I would say engineering incompetence, however there appears to have been no engineer, even of minimal experience, involved at all. Alignment sheets at best show only the location of pipe and components. They do not typically show any control interrelations, as for example as would have a set of PFD and PID drawings. In fact, alignment sheets do not even show the direction of flow. Those facts alone should have been obvious even to a most inexperienced pipeline engineer and that is something I find of grave concern in itself.
I also believe the practice to which NTSB refers of Massachusetts making an exception for utility work in the face of their typical P.E. seal requirements for other types of work, and often on facilities that present a far lesser danger to the general public, is extremely common. I am sure it is not only Massachusetts that allows such serious exceptions to public safety to go unhindered. Take a look at TEXAS. Which brings me to my first question. Do you plan to advise all the state engineering license boards of the NTSB recommendation that PE seals be required for utility planning?

Furthermore I note that the resulting pressure in the pipeline damaged in the corresponding accident investigation was at some 12 X operating pressure. That makes me wonder how a pressure regulator, apparently designed to open more when downstream pressure goes low, thus increasing downstream pressure basically unhindered, came to be installed into a system where a change to plastic material and apparently a lower MAOP, would most certainly result in overpressure of that downstream piping. Such an installation would normally have to also provide at least one safety feature, if not both of the following,
1) A maximum pressure override, closing an immediately upstream block valve, set to the lowest of MAOP of the lower rated pipe, or MOP of the downstream system,
2) At least one relief valve set to the same
I would have insisted on providing both. Obviously a regulator set to open, potentially increasing pressure to mainline levels, would not have provided a safe design to supply a low pressure distribution system.
I am very much pleased to finally see movement towards actually requiring professionalism in utility design and planning. Long overdue. Thank You


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[RVAmeche]

It's truly incredible that they didn't realize the regulator sensing lines were located in the header they were abandoning.

Do we know if there was a relief valve located downstream the regulator (hopefully sized for the max flow of the regulator) that didn't work for some reason? Or was the low pressure distribution system really only protected by the pressure regulator?

 

[davefitz]

I mostly agree with LittleInch, with some clarifications.

It is remotely possible that the physical / mechanical design of the revised piping system was reviewed by an engineer, but the procedure for upgrading the piping ( specifically the sequence of construction events) was not . The requirements for such upgrades that effect the public is that both the design AND construction procedures be reviewed by a PE.

The wording regarding utilies use a PE for all work might complicate the normal plant maintainance and repair procedures that occur daily 'within the fence" of a utilities' industrial facility is not likely to be accepted by the owners of large facilities that do not directly affect the public.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[Mbrooke]

What is MAOP? Sorry.

 

[LittleInch]

Maximum Allowable Operating Pressure.

Can be the same or lower than the design pressure but it the agreed limit of pressure that a pipe can sustain without risk of damage or permanent strain and to which protection devices should be present to prevent the operating pressure exceeding this pressure.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[charliealphabravo]

From my inbox this morning from NSPE...

I've got an exciting opportunity for you!

For more than 100 years, one of the most fundamental challenges faced by supporters of professional engineering licensure is the issue of state and territorial licensing exemptions in areas such as industry, government, and other areas of professional practice.

A pair of bills in the House and Senate (H.R. 2139 and S.1097) seek to the end the engineering license exemption for public utility pipelines. The bills are a federal response to the Merrimack Valley, Massachusetts incident that occurred in the fall of 2018, when over-pressurized gas pipelines exploded, destroying several homes and killing one person. This legislation is an important step in our fight towards ending licensing exemptions.

Express your support for H.R. 2139 and S. 1097.

NSPE's Government Relations staff has been hard at work to end engineering license exemptions, and we need your help to ensure this legislation goes through.

Ask your elected officials to support H.R. 2139 and S.1097.
In NSPE's conversations with congressional staffers, we have a lot more influence if we can show them their constituents care about and support this issue.

Contact your elected officials and ask them to cosponsor this legislation.

 

[1503-44]

Recovery effort

https://www.bu.edu/articles/2019/merrimack-valley-gas-explosions-joe-albanese/

 

[CWB1]

From the limited descriptions online it simply flabbergasts me that a system without a capable PRD at each residence and very limited shutoff capability was ever implemented in a major US city. Its easy to point fingers toward the first visible problem but ultimately the root cause of this disaster IMHO isn't lack of licensing or an incompetent employee as the NTSB reported but rather an unsafe design that likely was approved by multiple PEs in a corporation the size of Columbia as well as various regulatory agencies. Even more troubling is that Columbia's president also appears to have come up through their engineering dept and should've known better.

It appears that Mass has given oversight for the utility to a private engineering firm. Given the NTSB recommendations pushing bureaucracy rather than technical expertise and safety, it would be rather ironic if these PEs blessed a simple repair of the system without significant safety upgrades.

[link ]

https://www.bostonherald.com/2018/1...lumbia-gas-under-engineering-firms-oversight/

[/url]

 

[1503-44]

I agree with CWB1. It should be common sense for any engineer, that knows what they are doing, to install positive protection in the form of a relief valve immediately downstream of a pressure control device, when such device is connected to a system of lower maximum allowed operating pressure. My guess is that the key words are "know what they are doing". I would say that means NOT simply installing a pressure control valves, or a shutoff valve, any of which can easily fail, be disconnected, or have it's pressure sensing lines disconnected, as what happened in this Lawrence accident.)

A lot of the regulations, codes and standards mention "protection" of downstream lower pressure systems, however the only regulation that I can find which specifies exactly how these (distribution) systems (or any higher pressure system) must be connected to a lower pressure system is Alberta, Canada's "Technical Standards and Specification Manual for Gas Distribution Systems". Relief valves discharging to atmosphere at a safe location are specifically required by that technical standard. I uploaded a copy of that standard here, in case the online document disappears some day. I pointed the red arrows to the Relief Valves in the diagrams, as shown on page 33-37.

Protecting low pressure systems from high inlet pressure by discharging any overpressure to atmosphere, or in the case of liquids, an atmospheric tank, at a safe location is the way to stop killing our clients and blowing up their homes and businesses. Alberta got it exactly right in their standard. We all can get this right, P.E. or not, now that you have access to this document. Why PHMSA hasn't managed to get it into the US regulations, or why it is not required by ASME B31.3,4 or 8, or all the rest of the codes and regulations that purport to govern design of similar systems, is inexcusable ... IMO. Anyway, now we all know exactly what we should do. Thank you Alberta.