Bandwidth prices set to fall more rapidly in future

Brian Neilson

Submarine and terrestrial fibre-optic cable developments in Africa to explode over the next two years

 In the last decade or so, investment in Africa's ICT infrastructure has improved significantly. However, marked deficiencies persist in the backbone networks across the continent. Although countries on the African west and southern coasts have access to fibre connectivity through the SAT-3 undersea cable, an estimated 80% of Africa's international voice and data traffic is carried via satellite.

Following the glut of international submarine fibre-optic bandwidth that emerged around the turn of the century, the industry has shown signs of increased activity in the past few years. Since December 2006, several cable projects have been planned. From the US, two cables are being constructed to China, one to Australia and a further two to the Caribbean. Multiple cables are in various stages of development between East Africa and India.

In Africa, the effective high-speed internet services required for critical business, government and consumer applications have remained either unavailable or very expensive. Governments' awareness of this situation, and the perceived commercial attractiveness of the opportunity to close this gap, has given rise to the current frenetic activity for construction of submarine fibre cables on the continent.

There are at least ten submarine fibre cables and a significant number of supporting terrestrial cables that are either planned or under construction in Africa, with South Africa, Nigeria, Kenya, and to a lesser extent, Tanzania driving the demand for international connectivity. If all of these cables were to be constructed, companies and governments will spend in excess of US$6 billion on submarine and terrestrial infrastructure projects in Africa over the next twenty-four months.

This would create a significant capacity glut in some regions, despite the rapid growth in demand that can be expected. Considering only the cable systems planned for Southern and East Africa (including the critical market of South Africa, which contributes by far the largest demand in the sub-region), the total design capacity of all planned cables is 10 Tbps (up from the base of less than 400 Gbps today). Can this capacity be taken up? BMI-T forecasts that demand for this sub-region will be no higher than 120 Gbps by 2012 (up from 13 Gbps in 2006), growing to a maximum of 2,000 Gbps (2.0 Tbps) by 2019. This would represent a significant capacity glut in excess of 9 Tbps in five years time, and would remain a glut of 8 Tbps ten years from now - even if highly demand hungry projects like the Square Kilometer Array project are taken into consideration.

The most likely outflow of this seemingly obvious observation is that not all of the slated cable systems will in fact get off the ground. Furthermore, each of the cable systems that is launched, will light up fibres progressively, rather than commission the entire design capacity from Day 1. A possible scenario is shown below, assuming that each cable system is commissioned, and lights up 25% of its ultimate design capacity by the end of 2012. By 2015, a further capacity is brought on stream, bringing the total closer to 50% of design capacity.

Note that historical supply is based on total design capacity of the SAT-3/WASC/SAFE cable system, although it is not fully lit, and only a small proportion of the total capacity is availed in South Africa, the single largest market in the region. Future capacity growth is based on the assumption that only a portion of the design capacity of each of the planned cable systems is commissioned, and that each cable is lit progressively.

The good news arising from this analysis is that international wholesale prices for bandwidth will fall much more rapidly in future. This will remove one of the key inhibiting factors in the broadband internet provision in Africa, notably included the high cost of bandwidth and its monopoly provision. Importantly, competition is expected to increase dramatically in respect of the provision of international bandwidth in South Africa, with Neotel’s access to the SAT-3 cable landing stations and the advent of new under-sea cables due for launch, notably Seacom in June 2009. However, it is necessary to note that competition in national long distance and local access is also necessary, since the price reduction in international bandwidth will only be part of the overall cost structure. Real competition in these other legs of the supply chain will be equally essential if the overall rate of price reduction that many South Africans hope for is to materialise.