[kictanet] TEAMS and Open ended circuit

Fri Oct 9 07:29:24 EAT 2009

Aki,

a word from the economist ..

You cannot divide the cost of a system by the capacity one has ... You have to divide the cost by the capacity that can be sold ..

Let's assume we sell in the first year 10Gbps of the TEAMS capacity (jointly between all operators) and the investors want to have at least the return anybody expects from injected cash (I am taking the yields of the latest Infrastructure Bond as benchmark).

So:

130M USD to be returned in 5 years

1st year = 26 000 000 USD + 4% OaM = 31 200 000 USD

That makes per Mbps and month 260 USD

If we say we can sell 20Gbps then it would be 130 USD ... Let me stick to 10 for now ..

Cross Connection in Fujeira between the TEAMS Landing and the SMW4 Landing on lease for 4 x STM16 = 200 000 USD per month = 20 USD per Mbps and month

Onward Capacity from U.A.E. to London per STM16 approx. 150 000 USD per month = USD 60 per Mbps

Add USD 20 per Mbps for all peering and hosting charges in London and you will have: USD 360 per Mbps and month ...

This case assumes the 4 x STM16 are sold from day 1 in full .. Surely we can assume a straight line take-up that might peak in even 20Gbps after 12 month, then the first year will still be as above. The second year would be in the range of 210 - 230 USD .....

All the above is calculating 0 overheads at the operator's level .. 

No magic .. Just math ..

Rgds

Kai

   ----- Original Message ----- 
  From: aki 
  To: kai.wulff at kdn.co.ke 
  Cc: KICTAnet ICT Policy Discussions 
  Sent: Friday, October 09, 2009 6:55 AM
  Subject: Re: [kictanet] TEAMS and Open ended circuit

  Hi, its the techie and a non-stake holder. I'd like to add a contribution to the subject.

  Regarding the costs of internet dropping, a short summary. There is much more to the scenario than being let out by Service Providers.

  Consider Teams :

  - Mombasa to Fujiarah is a technically known as a Half Circuit. Half Circuits are cost reduction methods and have been in use in as early systems such as International Private Leased Circuits or IPLCs. The concept works very well and the Govt initiative got it right. However others may seem to exploit this scenario. 

  Assuming that the one time cost of the fiber cable between the 2 points ( Mombasa to Fujiarah ) was USD 130 million and the cable capacity was 40Gbits/sec, therefore the cost of 1 Mbit/s is about 3.25 USD on the half circuit. Because the cable is OWNED not LEASED on a monthly basis, Teams players get much cheaper cost bandwidth than compared to Seacom cable on the same route as it is leased.

  So the concept of cheaper bandwidth was spot on using a half circuit when the Govt initiative was done i.e that is how I understood it to be.

  Now comes forward connectivity. Because of the main landing stations at Fujiarah for other cable operators which are major transit inter-continental points for bulk purchasers like telcos who may even take up STM-16 capacities,  players on Teams may be able to get 10-25USD per 1 Mbit/sec. Bandwidth costs at these points has always been cheap, the problem was that no one wanted to bring that connectivity to this part of Africa. Seacom is a much longer and bigger project thus its costs for a 20 year commitment on the capacity is around 80-90USD Per 1 Mbit/Sec of bandwidth. ( IRUs )

  So once more, costs to Teams players is around 30USD per 1 Mbit of capacity on an IRU form Fujiarah onwards. Add the maintainence of the cable and costs go up. I'd estimate that the maximum cost is around 60USD per 1 Mbit/Sec of bandwidth for bulk purchasers.

  KDN was offering Seacom @600USD/Mbit as early as 1st week of August 2009, Safaricom had already done the move to Seacom within days of it going live and was offering similar via its other company OneCom. As soon as time become available, I shall do a simple analysis of the costs inland and list them here. Service Providers are  holding back due to other reasons.

  Solution :

  - I believe one major solution that is going to activate the price drops is for CCK to introduce a Virtual ISP license with an annual monthly fee of Kshs 10,000/-. VISPs do not own or operate any infrastructure. They would operate a customer support base and would be able to buy the cheaper bandwidth eg from KDN on Seacom/teams or Safaricom OneCom Seacom/teams and pass the reductions to the end users. I think this avenue should be looked into.

  Content: 

  - Content is something waiting to happen and there are many examples. As I mentioned on another list, take the example of Kenya National Museums and convert this into online catalogue content for use by Kenyan Schools and many others who use the internet. To create a virtual experience would be an achievement. If no one has done this within the next 24 months, then I hope some of us will get to it and move it to a newer level. ( Not advertising nor marketing here and its a techie thing, but some of us have setup a site like a notice board where we share what projects are in the pipeline : www.projectkenya.info )

  My amatuer thought, corrections are welcome.

  Rgds.

------------------------------------------------------------------------------

  _______________________________________________
  kictanet mailing list
  kictanet at lists.kictanet.or.ke
  http://lists.kictanet.or.ke/mailman/listinfo/kictanet

  This message was sent to: kai.wulff at kdn.co.ke
  Unsubscribe or change your options at http://lists.kictanet.or.ke/mailman/options/kictanet/kai.wulff%40kdn.co.ke
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.kictanet.or.ke/pipermail/kictanet/attachments/20091009/01a216e9/attachment.htm>