2014/05/07 Submarine cables in Sub-Saharan Africa: terrestrial networks need to keep up

Seven major new submarine cable systems have been completed in Sub-Saharan Africa since 2009, and several more are planned. This development has transformed the availability of bandwidth in coastal regions, and competition between cable operators is leading to a rapid reduction in the price-per-megabit-per-second to end users (for example, TeleGeography states that the median monthly charge for an STM-1 circuit from Johannesburg to London has fallen by more than 40% per year since 2009 – see Figure 1). This article reviews progress in delivering this new capacity into businesses and homes across Africa.

Figure 1: Median monthly recurring charge for STM-1 circuit between Johannesburg and London, December 2009–December 2013 [Source: TeleGeography, 2014]

Every African country has international fibre connectivity, but lack of competition at the national level is keeping prices high

The availability of abundant, inexpensive fibre capacity on the coasts of Africa is clearly a welcome development, but it is not sufficient to enable the provision of broadband services throughout the continent. For this, the continent will also need:

• international terrestrial links to bring connectivity to landlocked countries (of which there are 16 in the region)
• national fibre networks to link population centres to the international gateways
• local broadband networks to deliver 'last mile' connectivity.

Every country on the African continent apart from the Central African Republic, Eritrea and South Sudan has some form of fibre connectivity to one or more submarine cable landing stations and a large majority of countries have more than one national fibre network.1 However, many of these networks have been built by mobile operators for their own use, rather than for resale, while others are owned by national governments and utility companies that often lack the skills or the inclination to market them effectively to third-party ISPs and other commercial users.

According to our research, 35 of the 48 countries in Sub-Saharan Africa have no competition among national fibre providers (see Figure 2). Eight have limited competition – that is, two providers besides the mobile companies, usually the incumbent fixed-line operator and either the government or the electricity transmission company. Only five countries (Kenya, Nigeria, South Africa, Zambia and Zimbabwe) can be said to have effective competition among multiple players. In countries that lack effective competition, fibre connectivity in cities that are far removed from submarine cable landing stations often costs five or six times as much as it does at the landing station.

Figure 2: Degree of competition in domestic fibre markets in sub-Saharan Africa [Source: Analysys Mason, 2014] – Click to enlarge

Public–private partnerships can be used to develop national connectivity, but no single formula can be applied in every case

International fibre operators such as Liquid Telecom also point out that there is a 'chicken-and-egg' problem with low demand: because the cost of operating a long-distance fibre network is largely fixed, low demand tends to keep prices high, which in turn discourages take-up. Intervention by governments or international financial organisations can help in this situation. For example, the World Bank is assisting several public–private partnership (PPP) projects involving the construction of new submarine landing stations and/or high-quality long-distance fibre networks to try to improve connectivity at the international and national levels, and make it available to potential customers.2 In this context PPP can be defined as a financing mechanism or operating model involving a formal agreement between a government and a private investor/operator with the objective of developing, operating, maintaining or commercialising new connectivity. A wide range of ownership- and risk-sharing arrangements between the public and private partners can fit within this definition, as well as various forms of operating model.

Recent World Bank PPP initiatives have demonstrated that there is no single formula or approach that can be applied in every case. One possible model is that all sector operators unite to form a private company for the purpose of building, owning and operating the national backbone as a wholesale operator. In this approach, the government contributes a subsidy with no related ownership to ensure national coverage, including rural access points, open access, non-discrimination and cost-oriented pricing. Another possible model is to form a holding company to own the assets while another company is brought in to operate the network on an open-access basis – that is, selling access to anyone who wishes to purchase it. The separation of ownership and operation/sales helps to avoid underutilisation of the asset, but there is still a risk that, in the absence of real competition, the operating company may not really be motivated to reduce costs and maximise uptime.

Uptime is a major issue on terrestrial fibre networks in Africa. Common problems that lead to availability being much lower than in more-developed regions include:

• fibre not being buried deep enough (resulting in frequent physical damage, both accidental and deliberate)
• poor quality splicing (which reduces the throughput and can result in intermittent faults)
• poor maintenance of manholes (which can lead to flooding and cable damage)
• poor systems and processes for fault management (there are even anecdotal accounts of maintenance companies deliberately sabotaging cables to create work for themselves).

Overhead lines can also be damaged by vehicles hitting the support poles, trees falling on the cables, wildlife (particularly monkeys) and vandalism. Lines carried on high-tension electricity pylons are generally more reliable because the cables are carried high in the air, the pylons are generally situated away from roads and the proximity of high-voltage power lines discourages vandals.

Wananchi's Zuku network in Kenya demonstrates the potential for local fibre-based broadband services

Many large business customers in Africa now have access to local fibre networks (for example, Liquid Telecom's ZOL Zimbabwe network in Harare) and some cities have local-level competition. Consumers and SMEs are not so well served, mostly having to rely on 3G mobile networks or small local WiMAX operators (see Robert Schumann's article Consolidation threats and opportunities in Africa – from another perspective). It is generally accepted that most consumer broadband access in Sub-Saharan Africa will take place via mobile networks because of the small size and poor quality of local copper networks, which are used elsewhere to provide fixed residential broadband access.

Non-traditional players are building new local access networks in a few places. The Liquid Telecom's ZOL network covers some residential areas in Harare as well as business districts, but one of the largest projects is Wananchi's Zuku fibre network in Kenya. Zuku passes about 200 000 homes in Nairobi using a combination of HFC and GPON, and is expected to go live in Mombasa shortly. It may then be extended to other cities in East Africa.

Wananchi sells Zuku broadband as part of a triple-play package with pay-TV and fixed voice services for KES2999–9999 (USD36.5–118) per month, depending on the access speed and TV package, but the operator says that broadband is the component that the majority of customers value the most. Wananchi is using overhead cabling, mostly attached to electricity poles, which helps to keep construction costs down. Zuku is achieving take-up rates comparable to those of cable operators in many European countries.

If companies like Liquid and Wananchi can continue to make the economics of fixed broadband access work, then far more African consumers could gain access to the type of broadband connectivity that their counterparts in developed countries already enjoy.

Source: Analysys Mason