Google followed by Facebook are the behemoths of submarine cables with global investments in 22 cable systems over the last 10 years. It seems, their submarine cable investments will continue unabated for the foreseeable future. Expectedly hence, they are also the key driving force for technology innovations in wet plat and dry plant leading to many firsts in subsea data transmission. In parallel they are developing captive hyperscale datacenters in developed markets and pursuing in country presence in emerging markets in collaboration and co-investment with local providers. The objective evidently is to accelerate availability of affordable broadband. For Africa, it includes Kenya, Uganda, Ghana and Nigeria, where they have already ventured. In doing so they bring together learnings from scaling global infrastructure with local provider’s knowledge of the market landscape to develop new working models with open infrastructure.

International Internet Bandwidth Growth

10G IP Transit & Wavelength Prices

Africa going by market trends needs new submarine cables and the funds to develop them. It has the highest demand growth. It also has the highest price discrepancy. Based on weighted median pricing for 10G IP transit, its $0.6/meg in London, while in Johannesburg its 8X higher at $5/meg. This price differential ought to be narrowed significantly by the power of subsea connectivity. New submarine cables can enable a thriving business ecosystem at the intersection of connectivity, cloud, content and consumer, that will improve the wider socioeconomic landscape. Possibly the renewed interest from US to up investments in Africa will also play an important role.

Africa however is different, or rather unique, with 54 countries. Despite proven credentials with submarine cables and gigantic budgets to develop network infrastructure, when it comes to Africa, the landscape and challenges it presents are very different. Hence Google led Equiano and Facebook led Simba, the new submarine cables reportedly in planning and deployment, will have to overcome the challenges that are daunting but not unsurmountable, to script a unique success story. This article aims to foresee the challenges and options available to overcome the challenges.

Submarine cables Google and Facebook have developed or invested in till date are primarily point to point, between countries that have vibrant multi-provider competitive market landscape. Contrary to this in continental Africa, the cable system needs to connect multiple countries and every country has its unique challenges. Further none of the pan African telco carriers have CLS in more than 2-6 countries and multiple collaborative ventures need to be forged to achieve targeted country coverage. Hence it all boils down to – who to collaborate with and what should be the modality of collaboration, given Google and Facebook would look for carrier neutral cable landing.

Google announced work in progress for Equiano a week back, stretching from Portugal to South Africa with a branch landing in Nigeria. With the same subsea route as WACS, it will span nearly 14000 kms. That indicates it would be a cable system of the order of $400M. It will be fully funded by Google and that’s understandable, since none of the providers in Africa would be interested to co-invest $25-33M for a submarine fiber pair (FP). Further given it’s an SDM cable system that supposedly indicates it could be 16FP system, an industry-first, and a step ahead of DUNANT with 12FP. Conservatively with 120x100G or 12T/FP, it would mean design capacity of 192T. Comparing with DUNANT with 20T/FP, Equiano could be 320T. Featuring yet another industry first – fiber pair switching, Google has contracted ASN for turkey supply and deployment in Q4 2018. It will have 9 Branching Units (BU) positioned close to countries they intend to land Equiano in future. Given RFS for connecting South Africa is 2021 followed by branch to Nigeria, possibly in 2022, other countries will be targeted beyond 2022. Much is expected to change in the 9 countries in next 3-4 years in terms of network infrastructure, carrier neutral datacenters, burgeoning wireless broadband subscriber base, and above all content consumption.

How relevant is 16FP. Is it unjustifiably and even intimidatingly overengineered, possibly a show of unquestionable dominance! Doesn’t seem so if we take a closer look. The possibility with Equiano landing in 10 countries 5-years down the line with each country needing 5-10T Internet capacity is quite real and that would make multiple FP relevant. Some of the omnibus FPs could be the currency of remuneration to cable landing parties that land Equiano, as Google deals with the morbid question– what’s in it for us. It’s also possible that in-country providers with CLS, or a consortium of providers develop CLS ground up, pay for the submarine branch and get access to unprecedented international connectivity with attractive economics. Thus, they will not be dependent on suspicious funding hiding debt-laden approach to fulfill submarine connectivity aspirations. Further South Africa and Nigeria will need multiple FPs to be lit to cater to skyrocketing data consumption.

Who will land Equiano in Lisbon/Portugal, Cape Town/South Africa and Lagos/Nigeria! For Portugal incumbent operator Portugal Telecom and Tata Communications own and operate CLS. Given the anticipated need for cross-border terrestrial long haul to interconnect with DUNANT in Saint-Hilaire-de-Riez, France, it’s likely that Google will go with Portugal Telecom. For South Africa, SAT3 and WACS land in CLS owned by incumbent provider Telkom South Africa (TSA). MTN was in the fray to construct a new CLS near Cape Town a year back, reportedly to land ACE, as the last leg of 5000 kms was to be constructed from Sao Tome and Principle. However, any progress hasn’t been reported, unless they have decided to work in stealth mode and come with big bang RFS announcement. Still it seems Google is likely to join hands with MTN for cable landing for South Africa. Further MTN already owns and operates CLS in Lagos, Nigeria, and Accra, Ghana as the landing party for WACS. Thus, MTN can provide one stop shop cable landing in South Africa, Nigeria and Ghana.

While Equiano provides a new route from Portugal to South Africa, what about route protection. Backup on WACS not sustainable, as subsequent upgrades in WACS will progressively yield lesser capacity, amidst cacophony of consortium members. Further WACS and surely SAT3 cannot provide terabit-scale backup. Even if partial backup till Nigeria is considered, MainOneGlo1 and even ACE are unlikely to gear up for terabit scale backup. In this context Google is reportedly considering a 4 FP branch to St. Helena (Jamestown and Georgetown). This branch when considered in conjunction with the fact that Georgetown to Fortaleza is 2700 kms, leads to the possibility that Google is likely to plan for triangulation with Equiano, a branch to St Helena (Jamestown and Georgetown) that’s extended to Fortaleza, a span of nearly 7000 kms ($200M) and interconnect with MONET where they own Fiber Pair, to reach Boca Raton, Florida. Thereafter 1000 kms terrestrial long haul to interconnect with DUNANT, the third arm of the triangle, to reach France, and finally 1500 kms terrestrial backhaul to Lisbon, Portugal.

This heavy lift triangulation has a major flaw. The branch starting midway, somewhere close to Angola, to St Helena, will not protect connectivity for South Africa. Hence a vertex of the triangle must reside in Cape Town, South Africa. There are two potential cable systems, SAEx1 and SABR to reach Fortaleza and interconnect with MONET. Here SAEx1 is would be preferred as SABR lands in Recife, that’s 1000 kms from Fortaleza. SAEx1 and SEAx2, despite its jaw dropping $850M investment proposal for a subsea route that is yet to gain traction, has continued to stay put assertively on its value proposition to provide alternate subsea route from Singapore to Virginia Beach, with branch to Fortaleza. It remains to be seen if they can sign-up an OTT as anchor customer, with fiber pair ownership from Singapore to US, and leverage it to raise funds. SABR has also remained silent. Apparently Seaborn Networks that had planned to develop SABR may have lost interest and reportedly considering the option to abandon the submarine cable initiative.  

Facebook compared to Google has remained tight lipped so far and not come up with any official outlook on Simba. It’s speculated that the Simba cable system will encircle Africa. It’s not the first time that such an ambitious cable system has been envisaged. With route length of nearly 30,000 kms, it’s anticipated to be $800M initiative, twice the size of Equiano. It’s possibly less than what Facebooks invests for a captive datacenter. However it's not necessary that the three segments along West, East and Northern coast will be developed concurrently. For the purpose of complementarity with Equiano, the East and Northern segments may go for deployment up in the first phase. Alternately if Google and Facebook agree on fiber pair swap for both to gain protected route, then Facebook would also start with Simba-West. Its quite likely that Facebook will go with the same landing party as Google. Further they would prefer to interconnect with MAREA.

The selection of cable landing party in African countries will be the biggest challenge Google and Facebook must address. Google choose to keep it simple to start with, and land Equiano in Nigeria only. It remains to be seen if Facebook follows the suit or is more ambitious. With respect to pan African providers that could be considered for cable landing, the top of the stack has Orange, in West Africa. Their shareholding in incumbent providers in the West African countries facilitated ACE to land in 14 countries. Given Google has partnered with Orange to land Dunant in France, it remains to be seen, if partnership can happen for Africa. Further if Orange and Google agree to join hands, Facebook will not be able to or even want to join, as it would not want Simba to land in the same CLS that lands Equiano. This may gain credence if Google and Facebook agree mutually to swap fiber pairs, for subsea path protection. Hopefully Google, Facebook and Orange are in discussion to arrive at shared understanding.

A prominent submarine cable West Africa is MainOne, that owns and operates CLS in Nigeria, Ghana with Ivory Coast included in 2019, arguably the most important countries after South Africa. Facebook is more likely to join hands with MainOne, as they already collaborate to extend MainOne fiber footprint in Nigeria. MainOne could also facilitate cable landing in Cameroon vide their partnership with Camtel for Nigeria Cameroon Submarine Cable System (NCSCS) and in Senegal vide their partnership with Orange to land MainOne. A one-stop-shop arrangement with MainOne to land Simba in 5-countries isn’t a bad deal!

Another leading provider in Africa is MTN. While they are MNO in 6 West African countries, they only have CLS in Nigeria and Ghana that lands WACS. They could as long-term strategy join hands with Google or Facebook to construct CLS in multiple countries. This however is easier said than done. While Google or Facebook can bring funding, MTN has to get necessary license and deal with resistance from incumbent providers. It’s possible that if Google and Facebook throw their weight and position the cable system as enabler of digital economy that will attract investment and generate employment, the regulatory authorities would be amenable to help and facilitate the needful.   

One other provider in Africa is Vodacom. They are not known to have any CLS. It remains to be seen if they are ambitious enough to pursue OTT assisted business plans to construct new CLS. However, this requires them to have license in countries of their interest. It’s too much to expect they will acquire license to construct CLS and land submarine cable/s. Hence, they are expected to focus on South Africa, Mozambique, Tanzania and Kenya where they have fledging MNO footprint. Notably, SEACOM and EASSy/WIOCC already have CLS in these countries. Thus, it seems unlikely that Vodacom will develop the 3rd CLS in these countries to woo Facebook or Google.

Liquid Telecom pan-Africa terrestrial network

The dominant pan African provider is Liquid Telecom. They have a great terrestrial network spanning 70,000 kms covering 14 countries from Cape Town to Cairo, that’s now diversified with hitherto unavailable East-West connectivity reaching Senegal, Ivory Coast, Nigeria and DRC. They have had lukewarm interest in Submarine Cables for last few years, however something tangible is yet to show up. It remains to be seen if they are committed enough to pursue OTT assisted business plans to construct new CLS. However, they would face the same challenge as Vodacom, that SEACOM and WIOCC already have CLS in countries where Liquid is licensed service provider. However, if they offer to develop CLS architected to meet expectations of Google or Facebook and integrated with datacenters in South Africa and Kenya and terrestrial connectivity to landlocked countries, it could be a differentiated proposition.

Owning and operating a CLS is not sufficient prerequisite to land a new submarine cable. It needs to be assessed if the CLS is fit to land the submarine cable and remain operationally consistent for 25 years, given nearly all were constructed 10 years back. Hence provision for space, power and other facilities needs to be assessed. Even after the essential prerequisites for cable landing are met, Google and Facebook will insist on provider agnostic open access. Considerations for 25-year sustainability will lead to the decision – to land in existing CLS or get the chosen landing party to construct new CLS. Alternately let the landing party pay for the submarine branch from the trunk and be responsible for cable landing.

East Africa is better placed compared to West Africa when it comes to submarine cable landing. There are a smaller number of coastal countries - Kenya, Tanzania, Mozambique and Somalia, other than South Africa, that are sizably populous, riding high on mobile-first approach to most facets of digital economy, commerce, entertainment and even education. The leading providers with multiple CLS are SEACOM and EASSy/WIOCC. While three other cable systems are under construction – PEACEAfrica-1 and DARE1, however they haven’t been able to stir market interest and unlikely to be preferred by OTTs. Some industry observers even question if they will see the light of the day.

SEACOM owns and operates 4 CLS, in South Africa, Mozambique, Tanzania and Kenya. They have had a credible track record, diversified from wholesale to enterprise services, developed domestic fiber infrastructure in these countries and cross-border connectivity to neighbouring land-locked countries. They however don’t have retail play, except possibly FTTH ambitions. EASSy/WIOCC owns and operates 6 CLS, in South Africa, Mozambique, Madagascar, Tanzania, Kenya and Port Sudan. They too have had credible track record with wholesale and enterprise services. Either of them could be a good choice for Simba-East for one stop shop cable landing in 4/6 countries.

Finally, there are the well-known incumbent monopolies in Djibouti and Egypt. Djibouti is a subsea stopover post, point of capacity interconnect and also the entry route for East Africa to reach Ethiopia, Kenya and Somalia. In Egypt, Telecom Egypt is part of subsea folklore, for their Red-Med transit and pricing monopoly, that incidentally has improved considerably in recent times. That could slacken traction for alternate route through Israel, Jordan and Saudi Arabia, being explored by some of the new cable systems.

To conclude, most of the OTT-Telco collaborative possibilities explored in this article are possibly underway for last 12 months and even longer, deliberating win-win co-investment propositions. Having worked closely with more than 25 African Telco’s over the last 11 years, I understand these conversations take time to gain traction and conclude contractually enforceable arrangement. It would be interesting to track further developments with Equiano and Simba and witness market transforming partnerships emerge.

Will Equiano and Simba with their nearly inexhaustible capacity promote a submarine cable monopoly in Africa and prevent other countries from developing submarine cables. This has been the concern raised by some industry observers. However, Africa over the years has become united and assertive enough to protect interests of its countries and create avenues for each country to pursue their digital growth journey. On the other hand, these submarine cable systems will over the years develop a content and connectivity ecosystem that will impact the socioeconomic development. To conclude, Africa awaits Equiano and Simba to unleash the power of connectivity and open doors for progress and prosperity.