So it turns out there is a bit more I want to say about the Merrimack Valley Gas Explosions.
In yesterday’s post, I was rather critical of the immediate recommendations that were produced by the investigators. They seemed to be very focused on the human error side of things, ensuring that component people approve the work and filling in the correct paperwork. What I would have liked to have seen was an investigation into more fundamental engineering solutions.
When designing plants, assume everything will fail at some point. Make sure your design can handle it safely. So how were the regulators able to overpressure the system. Why wasn’t there something independent in place to either relieve the pressure or shutdown the system automatically?
Things will brake and people will make mistakes. We should do everything possible to ensure that things are well maintained and that people are in the best position to make the correct decisions. But relying on someone making the correct decision every time will eventually lead to someone having a very bad day. It is far better to have layers of protection in place so that when one component (either physical device or human decision) fails, something else is able to take over and protect the system.
To finish with I would like to point out what the operating company has done after the accident. Rather than attempting to inspect or repair the existing low pressure network, it has been completely replaced with a new high pressure network right up to everyone's home or business. Rather than having 14 centrally located regulators, the pressure is dropped at the gas meter as it enters the building. So no longer can a single incident overpressure the whole network1. Though now instead of the 14 regulators, there are thousands of them. But at least only one building will be blown up then. Not sure I am filled with confidence.
- well, the high pressure network is supplied from somewhere. Perhaps it could still get overpressured… ↩