A deal and its end

On the issues arising from the Antrix-Devas agreement and the gross procedural violations in the steps initiated to implement it.

COMING as it did soon after the revelations about the irregular auctioning of 2G (second generation cellular telephony) spectrum (900 MHz,1.8 GHz & 1.9 GHz) and its impact on the exchequer, the controversial commercial agreement between Antrix Corporation Ltd., the marketing wing of the Indian Space Research Organisation (ISRO), and Devas Multimedia Pvt. Ltd. on the lease of satellite capacity in a portion of the S-band allocated to ISRO for satellite services (2.4-2.8 GHz) has been viewed by the media through similar glasses. This has resulted in a public and policy perspective on the entire issue that misses out important aspects emerging from the contract signed on January 28, 2005, and its implementation.

Notwithstanding the February 17 decision of the Cabinet Committee on Security (CCS) to annul the contract, the issues arising from the deal and the gross procedural violations in the steps initiated to implement the contract need to be discussed to gain a proper perspective on the controversy. The Department of Space (DoS), which is the same as ISRO, had referred the case to the CCS on February 9 on the basis of a directive of the Space Commission on July 2, 2010, to cancel the deal in view of the national strategic and societal requirements that had arisen since signing the agreement. The urgency to effect this decision was in no small measure linked to the expose of the contract details in Business Line and The Hindu in early February.

Satellite communications operate over the radiofrequency range 400 MHz to 30 GHz of the electromagnetic spectrum, from ultra-high frequency (UHF) band through L, S, C, and so on, up to Ka-band, which are shared between space services and terrestrial services. The frequencies and power levels in different frequencies are coordinated to ensure that there is no interference across services. The International Telecommunications Union (ITU) coordinates the allocation of frequency bands for different services through the Wireless Planning and Coordination (WPC) wing of the Department of Telecommunications (DoT) and its National Frequency Allocation Plan (NFAP).

However, ITU regulations follow a use it or lose it policy, and the onus is on the country to which a particular bandwidth and orbital slot have been allocated, on the basis of its filing with the ITU, to fill the slot with an appropriate satellite within the specified time and start using the bandwidth for the specified service. Thus, if India fails to use a band allocated to it, it will have to forgo that band.

India began using the S-band initially for Broadcast Satellite Services (BSS) in the INSAT-1 & 2 series (for Doordarshan) and subsequently for Mobile Satellite Service (MSS) as well in the INSAT-3 series. Apparently, India has lost the use of L-band and part of the S-band (2.3 GHz) owing to lack of implementation of appropriate national systems in these bands. It now has only 190 MHz of the S-band (2.5-2.69 GHz) for its use. Of this, a bandwidth of 40 MHz has been farmed out for terrestrial wireless services, which was divided equally between Bharat Sanchar Nigam Ltd. (BSNL) and a private operator. So ISRO currently holds 80 MHz of S-band for BSS in three orbital locations (designated by the ITU for India) and 70 MHz for MSS in four orbital locations, but operable only from one location at a time because of interference issues.

SatCom Policy

Satellite communication in India is governed by the SatCom Policy of 1997, the guidelines for the implementation of which were issued on January 12, 2000. According to these regulations (ISRO press release of May 8, 2000), the INSAT system capacity will be made available to non-government (private) service providers on a commercial basis subject to its availability, and the DoS will allocate this capacity. The DoS may also build capacity in the INSAT system for private users on request on a commercial basis (emphasis added). For the use of spectrum allocated to it, the DoS leases transponders' carried on satellites to service providers (customers), both government and private. They, for example, include Doordarshan (C-band and S-band), DTH operators (Ku-band) and VSAT operators (Extended C- and Ku-band).

Private customers pay lease charges at the prevailing market price, which for a common 36 MHz bandwidth transponder is $1-2 million a year, depending on its power and other characteristics. (Transmitter-responder or transponder is a device aboard a satellite that receives, amplifies and retransmits a signal on a different frequency, and the total capacity of a satellite in a given frequency band is distributed over many transponders.)

The contract

The Antrix-Devas contract, together with the option (as allowed by the agreement) exercised by Devas in June 2007, provides for the leasing to Devas of 90 per cent of the space segment capacity on two geostationary satellites, GSAT-6 and GSAT-6A (the option exercised in 2007), for 12 years on a 24x7 hour a week basis, which would use up the entire 70 MHz of the S-band available with ISRO for MSS. Also, this leased capacity was to be on a non-pre-emptible' basis, which means this capacity cannot be used by any other user for any other purpose.

According to the contract, Antrix would lease 5 CxS (uplink/downlink) transponders, each of 8.1 MHz bandwidth, and 5 SxC transponders, each of 2.7 MHz bandwidth, on each satellite, which would be located at 83E. The agreement was, according to its preamble, to enable Devas offer a satellite digital multimedia broadcast (S-DMB) service, a new digital multimedia and information service, including but not limited to audio and video content and information and interactive services, across India that will be delivered via satellite and terrestrial systems via fixed, portable and mobile receivers, including mobile phones, mobile video/audio receivers for vehicles, etc. (Devas Services).

The agreement therefore provided for two custom-built satellites for supporting Devas Services, including new multimedia applications of MSS. It must be emphasised that the agreement was in conformity with the SatCom Policy, in particular the emphasised part above.

(Usage by Devas is 90 per cent because the transponders are actually of 9 MHz bandwidth, of which 8.1 MHz was earmarked for Devas and 0.9 MHz was for ISRO's own use. The C-band signals are feeder links for the earth station to communicate with the satellite and the S-band signals are for Devas Services. By simple counting it would seem that the total bandwidth in the two satellites exceeds 70 MHz, but there is some frequency reuse in the plan.)

Terrestrial question

The DoS note prepared for the CCS deliberations of February 17 says that the use of terrestrial services by Devas was to meet last-mile connectivity, say, in high-rise buildings (essentially all major urban areas) through WiMAX kind of wireless technology. However, says the note, such dispensation might not ensure a level-playing field for the other service providers using terrestrial spectrum [as against space segment spectrum] especially considering the significant demand for terrestrial S-band spectrum and the current trends in its price.

According to the DoS note, the department had initiated serious discussions for introduction of satellite-based multimedia services in the country in 2003, following which an MoU was signed with M/s Forge Advisors, a U.S. technology consultancy and venture capital company, in March 2003 to explore mutually beneficial opportunities in digital multimedia services. (K. Kasturirangan was Chairman, ISRO, at that time.)

According to Antrix sources, this was the only company that came up and made a viable presentation on using S-band for multimedia services. However, ISRO/Antrix did not announce any explicit window of opportunity for potential partners for developing digital multimedia services in the country. The last S-band transponders were flown aboard INSAT-3C in January 2002, which included two for BSS and one for MSS, and aboard GSAT-2, an experimental communications satellite, in May 2003, which had one for MSS. Since Doordarshan's switch over to Ku-band, effectively S-band had been vacated and if an appropriate platform was not launched soon India stood to lose this bandwidth, which, according to ISRO, was obtained after a lot of negotiations at the ITU.

Devas and after

In 2004, Forge Advisors founded Devas Multimedia Pvt. Ltd. in India. Significantly, according to the note, a joint venture was mooted initially but in December 2004 the Antrix board proposed to have a contract for leasing capacity on S-band satellites that ISRO was planning. Interestingly, the preamble of the contract says: Whereas Antrix and Devas understand and accept that they will collaborate to build, launch and operate satellite(s) and the Devas Services and recognise that enabling Devas Services and activities related thereto requires execution of interdependent technical and business activities (emphasis added). The preamble thus sets out the premise of the contract to be more than a mere leasing of transponders. To operationalise the agreement, Antrix had committed to developing and launching the two satellites, GSAT-6 (which has also been designated as INSAT-4E) and GSAT-6A.

Indeed, the 2,300-kg satellites are stated to incorporate several advanced technologies. They use high-power 100 Watt Travelling Wave Tube Amplifiers (TWTAs) for delivering five high-power spot beams in S-band (with frequency reuse) to cover the entire country. Notable among the technologies is the development of a 6.5 m unfurlable antenna (UFA).

For delivering any of the direct-to-user services of the kind Devas envisaged in the low-frequency bands L and S, a very large antenna in space is needed to concentrate the power sufficiently so that receivers with small antennas, such as portable or mobile phones, receive fairly strong signals.

During launch such an antenna needs to be stowed so that the satellite can sit in the launch vehicle shroud, and it is unfurled in space using complex structures. This, indeed, was a significant development by ISRO considering not many space agencies in the world have developed UFAs.

The GSAT-6 project, being custom-built for Devas' advanced technology applications, involved substantial R&D activity by ISRO. According to Antrix sources, the prolonged negotiations, from 2003 to 2005, were basically because ISRO itself was unwilling to take the entire risk of building satellites involving new and untested technologies. The contract became possible only after Devas agreed to bear the cost of the satellites as well.

Accordingly, the contract provided for Devas to be involved in design reviews as well as the periodic project reviews. It is claimed, which is not unreasonable, that such involvement was necessary to ensure that the satellite characteristics meet the functional specifications for the multimedia services that Devas would implement.

The satellite cost was built into the payments Devas would make through what has been termed as upfront capacity reservation fee (UCRF). The agreement provided for a UCRF of $20 million per satellite. In addition, the schedule of payments included annual lease charges, starting at $9 million a satellite (for 5 + 5 transponders), which was to be increased by 25 per cent when Devas became cash flow positive, during the operational life of the satellite as well as critical component acquisition fees (whose exact amount is not known) towards building the customised satellite.

Contestable calculation

According to the background note issued by ISRO on February 10, the total amount payable by Devas was $300 million (Rs.1,350 crore) over a period of 12 years. Devas has so far paid Rs.58.38 crore as the first instalment of UCRF for the two satellites (in June 2006 for GSAT-6 and in June 2007 for GSAT-6A). According to Antrix sources, the amount payable by Devas was arrived at by working backwards after including satellites' cost, launch cost, insurance cost and risk margin. But this is contestable and the DoS note to the CCS does admit that the payments would not have recovered the full costs incurred by ISRO.

Only a detailed calculation, including the cost of capital for the upfront investment made by DoS using public money, the investment in R&D for developing the UFA, and so on, can tell if the satellite costs borne by Devas result in a healthy internal rate of return for ISRO that covers risk margin and provides a profit margin. However, as an expert put it, the real problem in the contract lies here.

If Devas wanted the capacity to be available over the lifetime of the satellite, it could have procured it on a turnkey basis as other service providers (EADS, for example) have done. In such a contract, the entire cost of the satellite would have been paid at the time of delivery minus performance incentive payments (typically 5-7 per cent) over the satellite's lifetime.

Devas would, of course, also pay for satellite operations and maintenance on an annual basis in a separate service contract with ISRO. In case of an in-orbit failure, ISRO loses only the unpaid incentive payments. In the Antrix-Devas deal, however, in case of an in-orbit failure, Devas has no liability while ISRO has to bear the costs of a failed satellite built with public money.

In sum, it is a procurement deal disguised as a collaborative project-cum-lease agreement, with all the advantages in favour of Devas and all the disadvantages deferred recovery of investment made, uncovered risk of failure, etc. exclusively borne by ISRO.

Approval process violations

But it is in the approval process that there have been even more serious violations of procedures laid down for any project by ISRO. For any project approval, the proposal is first placed before the Space Commission, which includes the Cabinet Secretary and a senior official, of the rank of Secretary, of the Finance Ministry. On being cleared, it goes to the Cabinet for approval. In the case of the INSAT system of satellites, which has multiple users, the proposal goes to the INSAT Coordination Committee (ICC) to coordinate the allocation of transponders and frequencies and then to the Cabinet.

However, according to the DoS note to the CCS, involving the ICC, which includes other ministries, the DoT in particular, and which acted as a check in the entire project approval and implementation process, had been dispensed with since 2004 for inexplicable reasons.

This is not the first time a substantial capacity on an INSAT system satellite has been leased to a non-governmental customer and this is also not the first time a satellite is being custom-built. In 1999, ISRO leased 50 per cent of the capacity of INSAT-2E (nine C-band transponders) to an international consortium, INTELSAT, at $110 million, based on the market cost of leasing a transponder. India was also a member of the consortium with a 2.1 per cent share. In this case, the transponder characteristics, such as beam shape, were built to INTELSAT's specifications and INTELSAT was involved in project review meetings, and so on.

However, in 1999 the rules for implementing the SatCom Policy were not yet in place. So the lease had to go through the various governmental approvals, including those of the Space Commission and the Cabinet.

In fact, according to former ISRO scientists involved in the INSAT-2E project, the deal was cleared in 1995 itself before the project approval and the fact of the agreement was placed before both the Space Commission and the Cabinet.

In 2005, all the 12 Ku transponders aboard INSAT-4A were leased to TataSky for its DTH operations. In this case, given that ISRO was permitted to negotiate directly with the customer as per the SatCom Policy rules, approval of the lease itself at higher levels of the government was not considered necessary. And, most significantly, the ICC had been rendered defunct.

Why under wraps?

But the case of GSAT-6/6A is significantly different from the earlier cases and the manner in which the approvals were sought is extremely curious. The project itself is customised for the specific commercial objectives of Devas. But, strangely enough, the existence of the Antrix-Devas agreement was not disclosed both to the Space Commission and to the Cabinet at the time of project approvals in December 2005 and October 2009. In particular, if Devas was contracted to make deferred payments to meet the cost of the procured satellite, this fact should have been stated upfront to both the bodies. Why this was not done is indeed baffling. While this does not seem to be an oversight, the real reason is actually not clear.

The Cabinet had cleared the GSAT-6 proposal in December 2005 at a total cost of Rs.269 crore (including a foreign exchange component of Rs.102 crore) comprising Rs.170 crore for the satellite, Rs.34 crore for insurance and Rs.65 crore for pre-investment for critical components. The approval for GSAT-6A was obtained from the Space Commission in October 2009 at a cost of Rs.147 crore excluding the cost of insurance.

According to ISRO's background note, the launch cost for the two satellites was Rs.350 crore. But even in the gap of four years until GSAT-6A was cleared, the apex bodies were not informed of the specifics of the projects. In fact, according to the DoS note to the CCS, the Commission had noted in its meeting of July 2, 2010, that it was being apprised of the contract for the first time.

Significantly, the Secretary, DoS, or the ISRO chairman, is also the chair of the Antrix Board of Directors, the Space Commission and the ICC. G. Madhavan Nair was the ISRO Chairman between September 2003 and October 2009 and must be held responsible for the nature of the deal as well as the serious procedural lapses. It is interesting to note that GSAT-6A was cleared just before he remitted his office to K. Radhakrishnan, the present Secretary.

According to the DoS note, based on concerns regarding the severe penalty clauses for delayed delivery of the spacecraft and for performance shortfall and violation of ICC guidelines of non-exclusiveness' in leasing capacity, besides security concerns because of the contract enabling subleasing of capacity, Radhakrishnan constituted a one-man committee of B.N. Suresh, a former member of the Space Commission (between 2006 and 2008) and of the Antrix board (between 2006 and 2007) to look into the agreement.

Subleasing of capacity is itself not an issue as it has been done by INTELSAT; it subleased to various TV channel services the transponders it leased from INSAT-2E. The real serious issue is that the contract allows subleasing to be done without approvals from Antrix/ISRO. The contract also allows through a somewhat vaguely worded clause total off-loading of all its rights to a third party. Antrix/ISRO only need to be informed of such events within a time period.

The B.N. Suresh report was further examined by the DoS with regard to the contract's technical, commercial, managerial and financial aspects and from the perspective of the emerging strategic and societal uses of the S-band that is being leased out. Apparently, the DoT had stated in a note dated July 6, 2007: The spectrum planned by DoS for strategic use not to be shared with commercial application as in the case of M/s Devas Multimedia. Based on this, the DoS decided to apprise the Space Commission in its meeting on July 2, 2010, on the issues and seek its guidance.

It must, however, be pointed out that while the Space Commission might not have been apprised earlier in a formal sense, and though it made observations on the deal five and a half years later, it cannot be argued that key members of it were not aware of the contract. Some members of the Space Commission are also represented on the Antrix board. In particular, Member (Finance) would have been aware of the sanctioned cost of the project by the Cabinet as well as Devas' deferred payments to meet the satellite cost.

Similarly, the Director of the ISRO Satellite Centre (ISAC), who is in both the bodies and is responsible for building the satellite, would be fully aware that the satellite he is building is to the specifications of a commercial contract between Antrix and a private party. Why these members did not raise the issue at an appropriate time is equally baffling. Also intriguing is why the Space Commission never inquired into the non-functioning of the ICC. Clearly, the Commission's functioning had been lackadaisical and seriously wanting.

The Commission's observations after its deliberations on July 2, 2010, included the following:

In view of spectrum being a vital national resource and in view of the limited availability of S-band spectrum and its need for meeting the emerging strategic and societal requirements, there is an imminent need to preserve S-band spectrum;

though the agreement provided for Devas to hybrid (space plus terrestrial last mile) connectivity, the cost of spectrum had not been factored in the costing of the deal;

the agreement allowed the possibility of Devas moving towards 4G services in the process of accessing the terrestrial spectrum;

late delivery of the spacecraft and performance failures/service interruptions do not affect Devas' revenue;

taking into account the time value of (public) money put upfront by the government, the estimated revenue does not appear to compensate for the investment cost by the DoS, and the cost of capital and the revenue shown in the agreement had not taken into account the risk factor in developing the satellite and their in-orbit flawless operation for 12 years;

the ICC mechanism be revived and Antrix functioning be restructured to ensure diligence of financial, contractual and legal aspects.

Before the Space Commission deliberated on the matter, the DoS had sought the opinions of the Ministry of Law and Justice as well as the DoT on June 16, 2010. The latter's opinion tendered on July 28, 2010, appears to be completely contradictory to its earlier opinion as well as the Space Commission's observation.

It said: [T]errestrial component of the BSS frequencies should also be given similar treatment as in the case of 3G and Broadband Wireless Access spectrum with regard to pricing and auction mechanism should be adopted for fair spectrum allocation.

Such a move, in any case, would involve amendment of the SatCom Policy as well as a realistic pricing of the terrestrial use of BSS frequencies, which will be a function of the number of earth stations involved and projections on the user base for such an application. But since the DoT has never carried out such an exercise and the extant SatCom Policy guidelines do not prohibit an agreement of the Antrix-Devas kind, the DoS cannot be seriously faulted for entering into a lease arrangement on a cost-plus basis. This point also applies to the observation made by the Space Commission on costing.

Devas, as any private company would, maximised its advantages, and if ISRO was found to be complicit in this, much like insider trading, for whatever reason, the culpability must be fixed.

Now that the agreement is being annulled in view of strategic uses, including those of the defence services and paramilitary forces, and other societal applications based on the space segment of the S-band, it may be pointed out that the defence forces have been using every conceivable portion of the electromagnetic spectrum, right from HF, UHF and VHF to L-band, S-band, C-band, X-band and Ku-band. This has never precluded the use of this spectrum for commercial services and by other agencies of the government. In fact, for some operators to get 2G and WiMAX services off the ground, the Services had to vacate a part of the spectrum they were occupying.

So what are these special strategic applications that require the vacation of 70 MHz taken for commercial use when the rest of the 120 MHz of the S-band lies unused. The DoT, too, with its long-standing turf war with ISRO in a bid to grab the S-band, has lost the game to the so-called strategic applications. Also, what is not clear is how GSAT-6, which had been specifically built with its spot beams and UFA to cater to Devas' proprietary advanced technology on the ground, will be utilised to meet these newly identified national needs.

Clearly, something else is the issue. However, the end result is that from a situation of technology in search of a satellite, ISRO is left holding a satellite in search of a technology.