Setting the record straight on NBN Co’s 2nd satellite

NBN Co have made some recent announcements about increases to the wholesale data limits available on the SkyMuster satellite service. This is good news for those Australians located in remote areas unable to take advantage of other NBN technologies (ie. FTTx or Fixed Wireless).

Source : NBN Co

NBN Co is justifying this increase on the back of some recent re-engineering to bring into use the second SkyMuster satellite for this extra capacity rather than keep it as a “dormant backup” satellite.

In particular, the NBN Co press release of 27 June 2017 quotes Bill Morrow as follows:

“Late last year we made the decision to re-purpose our second satellite, previously slated as a dormant backup service to actively share the load in delivering more data to customers on the Sky Muster™ service. After spending the last year reviewing and testing the capabilities of the service, we are now comfortable that we have the capacity to offer increased data packages to retailers.” (Emphasis added)

As I was the CTO of NBN Co during the period in which the planning, design and procurement of the satellites for this services was undertaken I would like to put the record straight.

It was always NBN Co’s plan to use both satellites, once successfully launched and commissioned, to provide services to remotely located Australians in a load sharing configuration.

This was not the first time Mr Morrow publicly stated that the second satellite (launched in October 2016) had originally been “slated” as a backup satellite to be used only in the case of failure of the first NBN Co satellite (launched 12 months earlier in October 2015).

I was actually in the audience at the CommunicAsia conference in Singapore on 31 May 2016 when Mr Morrow made a similar statement. The SpaceNews website reported Mr Morrow’s speech as follows :

“That second satellite was originally intended to be there for reliability and redundancy,” Morrow said. “As the satellite people will know, the probability of something going awry after a certain period of time drops down to almost nothing. It doesn’t make much sense to have a $200-to- $300 million insurance policy up in the sky. So we have repurposed it for capacity.” (Emphasis added).

These statements imply that the original NBN plan involved spending a substantial amount of taxpayer’s money as insurance only for a catastrophic failure of the first satellite and that the current NBN team has just worked out that the second satellite can be put to good use.

The most obvious proof point is the fact that NBN Co from the start pursued multiple orbital locations (or “slots”)  in geosynchronous orbit from which to operate the two satellites. This was a hot political issue back in 2012 when the Shadow Minister of the time raised questions regarding NBN Co’s ability to secure the orbital slots.

In fact NBN Co pursued four orbital slots in order to ensure two suitable slots were obtained. In a response to Senate Estimates question on notice NBN Co clarified this in May 2012.

Source : Senate Estimates Broadband, Communications and the Digital Economy portfolio Budget Estimates May 2012 NBN Co Question on Notice No. 333 (http://www.aph.gov.au/~/media/Estimates/Live/ec_ctte/estimates/bud_1213/bcde/nbn_288-345.ashx)

If NBN Co was planning on using the 2nd satellite as a backup then only one orbital slot would have been required. Both the active and backup satellites would have occupied this one orbital slot. If the active satellite failed then the backup would have been in the right position to take over operations. Importantly all of the operating satellite dishes installed on rooftops around Australia would not have to be re-pointed to a different location in the sky.

But with two satellites in two separate orbital slots NBN Co would be able to use both satellites to deliver services in a load sharing configuration.

But further proof comes also from various public documents from both Quigley and Morrow NBN eras.

Firstly, there was the 2011-2015 NBN Corporate Plan which clearly states that “To meet this requirement [of 200,000 users], NBN Co will need to launch two 80Gbps Ka-Band … satellites”

Source : NBN Co 2011-2013 Corporate Plan released on 20 December 2010 page 71 (http://www.nbnco.com.au/content/dam/nbnco/documents/nbn-co-3-year-gbe-corporate-plan-final-17-dec-10.pdf)

Secondly there was NBN Co’s SAU submission to the ACCC with its original SAU in December 2011 which included a document on NBN Co Network Design Rules. This document provides an overview of the NBN Co’s technologies and includes the following in terms of the satellite design :

Source : NBN Co Network Design Rules (provided to the ACCC in support of NBN Co’s Special Access Undertaking dated 19 December 2011 page 27 (https://www.accc.gov.au/system/files/NBN%20Co%20Network%20Design%20Rules%20%28public%20version%29%20%2819%20December%202011%29.pdf)

There is no mention of the 2nd satellite being used as a “dormant backup” for the 1st satellite. Both satellites are required in order to provide the necessary capacity to users.

Thirdly, NBN Co engaged consultants from Boston Consulting Group to conduct an independent review of the Fixed Wireless and Satellite projects in 2014. After Bill Morrow joined NBN Co as CEO in April 2014, a report was issued on 7 May 2014 that covers covers the satellite project in significant detail.

The report confirms the need for 2 satellites to provide broadband services to regional users in a number of places. There is no mention of the 2nd satellite being deployed as a backup.

Source : NBN Co Fixed Wireless and Satellite Review – Final Report release 7 May 2014  page 26 (http://www.nbnco.com.au/content/dam/nbnco/documents/nbnco-fixed-wireless-and-satellite-review-07052014.pdf)

In  “Section 4.4 Proposed Actions”of the report recommends a delay to the launch of the 2nd satellite for operational reasons but highlights that “This timing [ie. delay] should also not impact the end-user rollout as the additional satellite capacity [of the 2nd satellite] should not be needed until this time given the speed with which the industry can connect end-users” [See page 40 of the report].

And if these statements were not proof enough, NBN Co commissioned the consultancy firm Ovum to do a global satellite comparison in April 2016. This report clearly states that NBN Co is using two satellites to deliver the capacity for 200,000 homes and businesses.

Source : Ovum Satellite Broadband : A Global Comparison 27 dated 28 April 2016 page 3  (http://www.nbnco.com.au/content/dam/nbnco2/documents/Satellite%20Broadband%20-%20A%20Global%20Comparison%20-%20FINAL.pdf)

The NBN Co Satellite program was always controversial and highly political. The orbital slots were a point of controversy as mentioned earlier. Private satellite operators that subsequently collapsed ran campaigns slamming NBN Co’s satellite engineers and overall program.

Despite all these distractions the NBN Co satellite’s have been launched successfully and are delivering what was always intended from two load-sharing satellites.

Let’s hope NBN Co are not continuing to play these political games by making misleading statements of the original purpose, designs and plans for NBN Co’s two SkyMuster satellites to be used to deliver broadband to remotely located Australians.

  • Darryl Boelcsfoeldi

    Love the Firstly, Secondly and Thirdly!
    NBN has been a political football since inception.. may have begun with the best intentions but how quickly it became a political step ladder for so many with an accidental by-product of providing a service to some of our regional brothers. What about FTTN…? We already had that, they were called Telephone Exchanges in the old days. Quite a good system as the copper cables to most locations were also located there. Anyway, Malcolm and the gang from the good ideas factory pat themselves on the back for a job well done. No wonder we have a government with a knife edge majority. Only mandate they have is to hang on by the skin of their well flossed teeth for as long as possible. Cheers !

  • Daniel Rossi

    600ms latency == dropped packets , timeouts and completely unusable for internet let alone business.