web analytics

UFB progress q1 2020

Crown Infrastructure Partners reports a total of 966,773 homes and businesses were connected to the UFB fibre network at the end of March.

A further 17,038 connections were added in the first three months of 2020.

Across the country UFB uptake is now at 58 percent. Rolleston remains the most connected town with a 76 percent uptake.

The extended UFB build, that’s the first and second phases, has reached 91 percent and is running slightly ahead of schedule.

Faster and faster UFB

There is a clear move to the fastest UFB plans. CIP CEO Graham Mitchell says just under 22,000 homes and businesses moved to gigabit connections in the first quarter.

Around 82 percent of New Zealanders can now connect to the fibre network. It operated 169 cities and towns across the country.

Outside of urban areas the Rural Broadband Initiative increased its footprint by 3,078 homes and businesses in the first quarter of 2020. The network also connected 52 marae — if you are an overseas reader that’s a Māori meeting house.

A total of 27 new rural cellular towers began operating in the quarter. There are now 86 new rural towers nationwide serving around 46,000 connections.

How are we doing?

The big picture is positive. The fibre network proved its worth before the Covid–19 lockdown, but that only went to amplify its importance.

Fibre reaches a little over four-fifths of New Zealanders. Things are a little less clearcut when it comes to the remainder of the country.

Those rural users who live in sight of an RBI tower and are not waiting for a spare connection have a good fixed wireless experience. Performance can slow a little at times, but for most purposes RBI fixed wireless delivers.

Likewise people with the good fortune to be serviced by a Wisp — wireless internet service provider — are likely to get the broadband they need.

The problem comes for everyone else. That’s, perhaps as much as 10 percent of the population, probably more like six or seven percent. These are people who don’t have fibre, a capable Wisp or line of sight to a cellular tower.

Given the small number and the importance of decent broadband, it should be possible over time to fill in more and more of these gaps. The cost per connection might be high and it may be unrealistic to expect these people to pay the same as city folk. That is a decision for telcos and government policymakers.

Shenandoah RCG tower

During the Covid–19 lockdown the Rural Connectivity Group switched on 29 new sites. The new sites connect about 2000 homes and businesses to fixed wireless broadband and increase the reach of the mobile networks. This includes mobile phone coverage for a further 110 km of state highway.

The RCG’s recent burst of activity means there are now a total of 92 sites.

RCG sites are part of the second phase of the government’s Rural Broadband Initiative.

Joint venture

The RCG is a joint venture between New Zealand’s three mobile carriers: Spark, Vodafone and 2degrees. It builds 4G towers in parts of the country that might not otherwise get cellular coverage. Another part of its job is to reduce the number of mobile dead zones on main roads.

RCG - rural connectivity group cell tower

RCG partner companies share the infrastructure. Towers have one set of access hardware and one set of antennae. Spark, Vodafone and 2degrees all offer their services from each tower. The towers are also open for wireless internet service providers or Wisps to use.

Rural Connectivity Group footprint increased by one third

Much of the work for the new towers was done before the nation went into lockdown. Nevertheless, increasing the footprint by a third at this time is an achievement.

A post on the RCG website quotes RCG chief executive John Proctor who says:

“At the time of the lockdown we had a large number of sites around New Zealand close to completion which we needed to secure and walk away from. Being granted Essential Service status means our team has focused on getting built sites ‘on air’ providing immediate service to as many rural communities as possible.”

Eventually the Rural Connectivity Group will run at least 400 towers. It will deliver broadband to around 30,000 homes and businesses in more remote parts of the country. Those numbers could rise if government finds more funds to fill in coverage gaps.

We’ve all heard about the digital divide. About how disadvantaged New Zealanders don’t have the same access to broadband as their better off neighbours.

We also know about the rural-urban digital divide. People who live outside the UFB fibre areas often make do with second class broadband.

There’s another digital divide that you don’t often hear about. That’s the performance gulf between two classes of Rural Broadband Initiative (RBI) customers.

Fixed wireless broadband

Fixed wireless broadband is used to deliver connectivity to RBI customers. It may be second class when compared with fibre, but on a good day the service isn’t bad. When it works well, RBI customers can see speeds that are more than adequate for most applications, including streaming high resolution video.

The problem is that some people, the people on the wrong side of New Zealand’s third digital divide don’t ever have good broadband days.

It’s a complex topic, so while engineering types may disagree with my explanation, here’s the problem described for non-experts.

Vodafone and Spark mainly use 4G mobile technology for the first generation RBI towers. This is often called RBI1.

Carriers have two sets of spectrum. The 700 MHz band is used to connect people who live further away from towers. The 700 MHz signals reach further, in some cases over 20km from a tower and do a better job of dealing with obstacles. You don’t need line of sight.

Urban standard, but in the bush

RBI towers often also include antennae using, for the want of a better term, ‘urban 4G spectrum’. This tends to work only where the customer can actually see the tower. Or as they say in wireless communications circles, where there is “line of sight”.

In effect the urban spectrum delivers much the same fixed wireless broadband performance that customers will get if they live in a city or town. It tends to be faster and is more reliable, which means fewer drop outs.

There are nuances. Wireless performance depends on a number of factors; distance, the nature of obstacles and, in some cases, atmospheric conditions, can all play a part. Some more remote RBI1 customers are lucky. Others are not.

Customers connected to the same tower can see a huge difference in performance. While some may be able to enjoy Netflix, others may struggle to get their email out.

Overload

A second, related problem, is that carriers engineered and managed their RBI towers to cope with the levels of demand they saw before the Covid-19 lockdown. This includes limiting the number of connections on a tower so that users get a reasonable experience.

Fixed wireless bandwidth is shared. When too many customers connect at the same time there isn’t enough to go around. It requires a careful balancing act. The calculations used to manage numbers date from before the lockdown. Since the lockdown all broadband demand has increased dramatically. Which means towers that were coping before are now under pressure.

We don’t hear much discussion about the gulf between different classes of RBI1 users connected to the same tower, but the problems are real. Fixing it won’t be easy.

Last week engineers completed the first UFB stage. The so-called UFB1 fibre network reaches three quarters of the country.

UFB2 will stretch that to around 87 percent. We can take fibre further, but that needs taxpayer money. A lot of it.

When New Zealand built its copper telephone network, government saw it as a nation-building exercise. Copper phone wires reached almost everywhere.

The number you often see quoted is that it reached 99 percent of the country. It could have been one or two percent less. That’s not the point.

Copper went everywhere

What’s important is that it felt as if copper reached every part of New Zealand. Perception is important.

There’s no technical reason the fibre network couldn’t do the same. The arguments against running fibre everywhere are economic. A nationwide fibre network is expensive.

Yes, it was expensive laying copper to outlying settlements and buildings. We did that at a time when there was less money around.

State-owned monopoly

We also did it at a time the telecommunications network was a government owned monopoly.

The copper network was built as a public service, not a profit making business. Laying copper to the nation’s furthest reaches and maintaining the network created good-paying jobs for workers in regional New Zealand. That would have been a consideration. We rarely hear that argument today.

In a sense it was still about getting the maximum return on the investment, but not in the way modern companies measure investments and returns. There was a social component.

How far can we go with fibre?

We’re not about to go back to a state-owned telecommunications monopoly1. But there is still a social component to network building. So how far can we go given today’s conditions?

The easy answer is somewhere between the 87 percent already earmarked and the 99 percent the copper network achieved. It won’t be 99 percent, it will be more than 87 percent.

If pushed I’d say a little over 90 percent in the next five years with further add-ons later. But that depends on many moving parts. It also depends on technology not changing, which experience says is a mug’s bet.

Brutal economics

Many forces drive network extension decision making. The most brutal economic fact is that the further you go, the more it costs to add each extra address to the network.

By the time you get to the last few percent the cost is way higher than can be justified by an investor looking for a rational economic return. At least as things stand today.

A nation building government could find the money.

The good news is that fibre uptake is much higher than anticipated at the start of the UFB project. It’s already close to 60 percent and will climb well beyond that number.

This means investing money connecting what were once marginal addresses is now more likely to pay off.

There will be places not included in the 87 percent covered by UFB1 and UFB2 where connection makes sound economic sense.

Politics of fibre

Another force pushing the number higher is political. People in rural areas see people in towns getting Netflix and high quality streaming Rugby pictures. Their kids want to play Xbox games.

People want fibre and may pressure politicians to deliver. Never underestimate rural New Zealand’s ability to lobby government.

By now the people connected to fixed wireless broadband on the RBI network know they have second rate broadband. It will take a long time for their service to improve, if ever. There are stories of capacity problems.

Not everyone who wants a wireless connection can get one. It is unlikely rural fixed wireless will ever match fibre. That’s more pressure.

One way or another government needs to subsidise further network extension. So the answer to the how far will the network goes question is a matter of the willingness of governments and taxpayers to put people in rural New Zealand on an equal digital footing.

Before you ask how far will fibre go, ask yourself how much you are willing to pay?


  1. Discuss this by all means. Even if you think it is desirable, it’s unlikely. ↩︎

Clare Curran says: “We have to do more than better connectivity”. The minister for broadcasting, communications and digital media was speaking about the digital divide at last month’s Tuanz Rural Connectivity Symposium in Wellington.

Curran names ‘closing the digital divide by 2020‘ as one of her two big goals. The other is for technology to be the second biggest contributor to GDP by 2025.

Both are fine goals. Neither will be easy.

Rural digital divide

There are at least two types of digital divide. The first is geographic. To use Curran’s words, that issue is one of better connectivity. People in rural areas don’t have the same easy access to communications networks as people who live in towns.

Blame economics. The cost of getting fibre to a city dweller is lower than the cost of connecting someone living in a remote area.

Both types of customer pay the same price for a connection if they are on the regulated UFB fibre network. This means people living in easy-to-connect areas subsidise those elsewhere. Almost no-one complains about this subsidy. It’s a step towards bridging the rural digital divide.

And anyway, a flat rate simplifies billing for service providers. Billing is expensive, so simple bills help keep costs down.

Clare Curran at Tuanz rural connectivity symposium - digital divide

Drawing the dividing line

By the time the UFB project completes, 87 percent of the population will be able to connect to fibre. while the other 13 percent are more at risk of being the wrong side of the digital divide, they won’t all have second rate connections.

While the 87 percent cut-off point seems arbitrary, it is a reasonable place to draw a line. At least for now. Beyond that number each extra fibre connection gets more expensive to build.

Theory says that some point fibre isn’t economic. It’s not clear where that point is. When our ancestors built the copper phone network they managed to cover 99 percent of the population. We weren’t richer in those days, if anything the job was harder. So the choice about where to draw the line is as much about social priorities and politics as economics.

Beyond 87 percent

One day we may stretch the network further than 87 percent. There are already plans for still more fibre. Getting to 90 percent coverage wouldn’t be economic unreasonable. Getting to 100 percent would be.

As things stand there are more cost-effective ways of reaching the most remote 13 percent. Most involve wireless. That’s the approach favoured by the government subsidised Rural Broadband Initiative.

The problem is that wireless technologies are not as good as fibre. They offer slower speeds, are contested and they not as reliable. They are, in theory at least, cheaper. It costs less to beam radio waves across paddocks than to build fibre lines over them.

Contested

Contested means that users on a wireless network share bandwidth. If a lot of people are online at the same time everyone’s speed can drop. In contrast UFB fibre is uncontested. Contracts between fibre companies and the government guarantee performance.

Another problem with wireless is there is less network capacity. To get around this service providers impose data caps on users. Most fibre connections have uncapped plans. Wireless users get a set amount of data each month.

Although some fixed wireless data plans are generous, life is not carefree when you have to limit, say, your television viewing towards the end of the month to be sure of having enough data left for other uses.

Rugby World Cup

These issues could come to the fore during next year’s Rugby World Cup. Spark and TVNZ won the broadcast rights. Spark intends to stream games, the technology is like Netflix. We love the game nationwide, but Rugby’s heartland is rural New Zealand. Will fixed wireless networks cope when every connection on a tower is streaming high-definition television? Spark doesn’t say so in public, but some insiders have voiced fears about how this might go.

Wireless plans often cost as much as fibre plans. They offer less. Not a lot less. Yet on a like-for like basis they are more expensive than fibre plans. The extra cost may be an annoyance, but it doesn’t put a customer on the wrong side of the digital divide.

There’s a handy proof for this. Spark offers fixed wireless to customers everywhere on its network: rural and urban. Thousands of city dwellers have chosen fixed wireless.

If fixed wireless was dreadful you’d hear more about it. There would be a lot of angry people. Sure, there are some unhappy fixed wireless broadband customers. Yet citizens aren’t marching on Spark’s headquarters with pitchforks and burning torches. For many people it’s not bad.

Fixed wireless broadband may be inferior to fibre, but people who have it are on the right side of the digital divide.

Not there yet on rural digital divide

It isn’t quite that simple. There’s a limit to the number of connections a wireless tower can accommodate. This means its possible there are some rural users who can’t get a connection because their local tower is full. Carriers can add capacity, but it may not happen immediately. A handful of people may miss out.

A bigger issue is that fixed wireless broadband doesn’t reach all the last 13 percent of the population. Not yet. The exact number is hard to gauge. At the Rural Connectivity Symposium, someone said there could be as many as 100,000 homes still out of reach of RBI. That’s about 5 percent of total connections. I’m afraid I didn’t make a note of who said this.

Moving goal posts

There’s another aspect to this. A decade ago when the Rural Broadband Initiative was being set up, the aim was 5Mbps. That’s enough for web surfing, email and movie downloads. Today’s acceptable broadband threshold is the 30 to 40Mbps needed to stream HD video. RBI towers can and do deliver these speeds. Wireless internet services providers do a terrific job getting connectivity to remote places.

Today’s rural network performance is way past the 2009 test of acceptable broadband. Also, thanks to the wisps, today’s broadband network reaches further into valleys and outlying areas than the 2009 RBI architects expected.

Yet the question mark hanging over the Rugby World Cup tells us there is still a rural-urban divide. Today the bar isn’t 5Mbps or even 40Mbps. It’s “is there enough broadband for people in rural New Zealand to enjoy the Rugby World Cup on an HD screen?”

And that’s the rub. The urban-rural digital divide is a moving target. Some rural New Zealanders will always feel they have second-rate broadband right up until the fibre network reaches them. Whether that’s reasonable or economic is a political matter, not one for the industry. Are we as a country willing to spend what it takes to get fibre as far as we managed to spread those copper lines?