NBN Co quantifies how many HFC dropouts in a day is too many

NBN Co has quietly brought HFC under the same performance standard as fibre-to-the-node (FTTN), such that four or more “unexpected dropouts” in a 24-hour period would qualify a service to be investigated.

The company made the change to its operations manual [pdf] on February 22 to a section called PI - performance incident - thresholds.

Performance incidents (PIs) are a relatively recent addition to NBN Co’s operations; although iTnews was unable to determine exactly when they first came into force, the section is not in an operations manual that was effective as of July 2018.

A November 2021 version of the operations manual showed that PIs were for FTTN only; HFC - and indeed all other access technologies - were specifically excluded from the standard at this point.

NBN Co defines a PI to have occurred when there are “four to nine unexpected dropouts within one calendar day, occurring in the current day or any of the two previous calendar days.”

It defines unexpected dropout as a “temporary loss of connectivity”, but the definition excludes formal outages to the network.

In other words, a PI applies only to problems experienced on a single line, not when there is a wider issue impacting multiple lines or services.

It’s not clear why NBN Co suddenly decided to bring HFC under the PI threshold regime.

An NBN Co spokesperson would say only that it came as a result of an unspecified “consultation”, and did not respond to questions from iTnews about the reasons for the change.

“On February 22, NBN Co introduced a new performance incident framework to assist with management of service dropout issues on the HFC network,” the spokesperson said.

“This followed consultation with industry, and going forward will assist in diagnosing issues, raising service restoration tickets and determining when to dispatch technicians to customers’ premises.

“These changes better align our performance incident management of the HFC network to that on the NBN FTTN network.”

The PI thresholds still formally exclude all other access technologies, making the apparent urgency to bring HFC under the standard more unclear.

It may, however, be a move designed to quell longstanding problems with unexplained dropouts on the HFC network, and put a specific definition around what it would take to get a truck roll to examine the connection and try to remediate it.

Every retail service provider’s online forum has multiple customers’ posts that contain extensive logs of multiple HFC dropouts in a single day, and complaints of little change, even after technician visits.

Launtel CEO Damian Ivereigh last year called HFC “a complete dog’s breakfast” and said it “has got to be the worst product NBN [Co] have.”

“It creates so much work for us as an RSP, let alone the poor service delivered to our customers,” he said, adding that “the service drops out a lot, for some customers multiple times a day.”

“The NTD [network termination device] lights start flashing and your service is down until it can resync with the HFC node,” he wrote.

Giving HFC a written standard on the number of acceptable dropouts in a 24-hour period may be an improvement over the previous lack of a standard altogether.

However, most in the industry indicated to iTnews that a minimum of four “unexpected dropouts” in 24 hours is still too high, and that customers often complain after just one or two.

"Because work-from-home is so prevalent, the minute you get a dropout, people are logging faults,” one telco industry source said.

Under the PI threshold, that would still not automatically qualify a HFC service to be checked out by a technician, meaning that users and RSPs are likely to remain frustrated, with unstable internet and without recourse for action until it gets demonstrably worse.