COVER STORY

The promise of INSAT-2E

INSAT-2E could be a forerunner to the next series of India's satellites that will mark the beginning of an era of long-life mission satellites as the Indian Space Research Organisation enters the new millennium.

R. RAMACHANDRAN
recently in Kourou and in Hassan

THE fifth satellite of the second generation Indian National Satellite series, INSAT-2E, was well (as on April 5) on course to being parked in its slot at 83° E in the geostationary orbit (GSO) 36,000 km above the equator. It was launched aboard flight V117 of the European launcher Ariane 42P from the French space centre at Kourou in French Guiana at 3-33 a.m. Indian time on April 3. Ariane 42P denotes an Ariane 4 configuration with two solid strap-on boosters. As the Chairman of the Indian Space Research Organisation (ISRO), Dr. K. Kasturirangan, said in Kourou, it was a "textbook launch".

The French launch contractor Arianespace has an enviable track record in launch reliability of over 95 per cent. No wonder then that ISRO prefers to launch its communication satellites on Ariane rather than any other launcher system. INSAT-2E is the sixth INSAT satellite to be launched by Arianespace, and the next satellite, INSAT-3B, which will be the first of ISRO's INSAT-3 series, is also scheduled to go on Ariane later this year. In all likelihood, INSAT-3B will be launched by the new European giant launcher Ariane 5.

Unlike earlier INSAT launches, which were all shared ones with other satellites, V117 was dedicated to INSAT-2E. A dedicated launch results in better orbit injection parameters for a satellite. This, in turn, implies a saving of the on-board propellant and a consequent increase in the life of the satellite from its planned 12 years by at least four months. An increase in INSAT-2E's life also means higher revenue from nine of the satellite's 17 transponders. These transponders have been leased to INTELSAT, the international inter-governmental space consortium created in association with 133 countries. The lease arrangement fetches $10 million a year under an agreement signed on January 30, 1995 between ISRO and INTELSAT. Actually, thanks to a higher 72 MHz bandwidth of two of the leased transponders effectively, INTELSAT gets the capacity of 11 transponders of the nominal 36 MHz bandwidth. (India holds a 2.1 per cent share in INTELSAT and is represented on the board of the company by Videsh Sanchar Nigam Limited.) Given the fact that more than half of INSAT-2E's transponder capacity is being leased to INTELSAT under a contractual commitment, the significance of INSAT-2E for India is not so much in the augmentation of telecommunication and television channel capacity that will become available through the remaining eight normal C-band transponders. These will be shared between the Department of Telecom-munications (DoT) and Doordarshan. Doordarshan may acquire a larger share of the transponders. Even after INSAT-2E, there will be a substantial level of unmet demand for transponder capacity for the telecommunications sector.

Among the satellites of the INSAT system, INSAT-1D, launched in 1990, is well past its nominal life and is already in an inclined orbit, the inclination amounting to more than 1°. This handicap necessitates continuous tracking by ground systems, which service operators prefer not to undertake for reasons of reliability. Save some internal communication uses of ISRO, INSAT-ID's transponder capacity is, therefore, not available for any operational services. In fact, after INSAT-2E is nudged into its slot, it will be co-located with INSAT-1D and, in order to avoid interference, 1D's transponders will be shut off. To that extent, even if someone wishes to access its inclined orbit, its normal C-band transponder capacity will not be available. As for the rest several unforeseen satellite mishaps in orbit have led to a considerable shortfall in the availability of transponder capacity.

For instance, INSAT-2A is only notionally operational and its transponders are not available to service operators. A problem with its propellant tank design led to a depletion of the on-board propellant. The normal station-keeping operations could not be performed and the satellite has been allowed to drift into an unusable inclination. This satellite too is now being used only for ISRO's internal needs. The final blow came in 1997 when loss of power led to the abandonment of INSAT-2D. Besides all these problems, many of the solid state power amplifiers (SSPA) in the transponders, based on devices known as GaAsFETs, have failed in space, bringing down transponder availability in overall terms.

But for all these problems, the four INSAT-2 satellites, 2A, 2B, 2C and 2D, together should have nominally provided a space segment capacity of 86 transponders. The present transponder availability in the INSAT system is just 59. This includes the 26 transponders of Arabsat-1C, the replacement satellite for INSAT-2D. This satellite was purchased by ISRO with the payment received from the insurance company for the loss of 2D, moved to 55°E and renamed INSAT-2DT.

While INSAT-2DT's capacity more than compensates for INSAT-2D's loss in the C-band (used mostly for telephony), the shortfall created by the non-availability of 2A as well has led to ISRO taking on lease an additional 7 C-band transponders from the Thai satellite Thaicom. The non-availability of the entire capacity of INSAT-2E will have an impact on telecom service projections.

Indeed, a crucial component of the growing telecom demand, particularly in VSAT (very small aperture terminal) communication networks for business and emerging users such as Internet Service Providers (ISPs), will remain unaddressed till INSAT-3B goes up. This is owing to the absence of upper extended C (UXC) and Ku bands in INSAT-2E. Extended C-band is where all the existing VSAT services have been given frequency space to operate.

AP
The European launcher Ariane 42P blasts off from Kourou in French Guiana on April 3 to deploy INSAT-2E. The multi-purpose satellite is perhaps the most advanced in the INSAT series so far in terms of technologies used in the hardware and in the payloads.

INSAT is the only satellite system which carries this frequency band after it was released for VSAT use by the International Telecommunication Union (ITU) and it has been adopted for VSAT communications. INSAT-2D was intended to cater to the growing VSAT networks' space segment needs but its unforeseen failure and subsequent abandonment has created a critical situation for this application.

Moreover, neither Arabsat-1C nor Thaicom has the UXC-band that has been adopted by the INSAT system for VSAT communications. In fact, more than its impact on domestic telecommunications, INSAT-2E will mark ISRO's bid to enter into the international satellite market through its transponder lease arrangement with INTELSAT.

From the domestic perspective, the critical importance of INSAT-2E lies in the meteorological component of its payload. At present there are three INSAT satellites in orbit which carry the meteorological payload - INSAT-1D, INSAT-2A and INSAT-2B. But the infrared (IR) segment of the meteorological payloads in both INSAT-2A and 2B have been rendered non-functional owing to some unforeseen problems in the electrical wiring of this part, resulting in their burnt-out.

INSAT-1D already in a fairly inclined orbit, will soon become unavailable though its meteorological payload will still be operational. As 1D comes to the end of its life, INSAT-2E will become very valuable for the Indian Meteorological Department (IMD) for its services including weather forecasting and cyclone warning.

BUT before all these can happen, the launched satellite needs to be guided from its currently elliptical Geostationary Transfer Orbit (GTO), into which it was injected as it separated from the launcher after about 21.5 minutes of lift-off from Kourou, to the designated orbital slot in the circular GSO by a series of orbit raising and orbit trimming manoeuvres. The perigee (the nearest point from the earth) at injection was 250 km and apogee (the farthest point) 36,155 km. After the launch, the first signal from the satellite at ISRO's Master Control Facility at Hassan in Karnataka was received as scheduled after about 30 minutes of lift-off when the satellite in its orbital path in the GTO came within the MCF's radio visibility. All commands from the ground and signals from the satellite indicated that the satellite's health was good.

Of course, on-orbit glitches and unforeseen events are the bane of all space missions and the INSAT programme has had a fair share of those. In the four-satellite INSAT-1 series, 1A was completely lost and 1C had to be abandoned while in the INSAT-2 series, while 2A, 2B and 2C have all had problems of various kinds, including loss of IR visibility and non-availability of transponders, and 2D was totally lost. It is the loss of INSAT-2D that led to a delay in the launching of 2E because a thorough diagnosis of the failure had to be carried out in order that appropriate contingency measures could be incorporated in the already qualified design of INSAT-2E. Otherwise, INSAT-2E was scheduled to be launched in September 1998 aboard Ariane-44L, as in the case of all earlier INSAT-2 launches, as a co-passenger with a satellite of General Electric (GE). Ariane-44L denotes a configuration with four liquid strap-on boosters: the boosters are needed to put a greater payload mass of a shared launch in respective orbits, as against a dedicated launch.

The orbit-raising manoeuvres basically consist of the crucial three firings of the on-board Liquid Apogee Motor (LAM). The first firing, LAMF-1, which is nominally carried out during the third apogee pass, was successfully carried out at 5-50 a.m. on April 4, about 26 hours after the satellite was injected into the GTO. However, the firing lasted only for 16 minutes as against the planned 75 minutes owing to an unexpected on-board computer glitch which sent the system into a 'safe mode' operation. In this mode all thrusting action gets stopped and this caused some concern. Subsequent checks revealed that the satellite's parameters were nominal and that the performance of the LAM during the shorter firing was good. The safe mode was initiated perhaps owing to some random anomaly (which could include problems with the computer processor itself). The satellite was brought back to normal mode by ground commands and was performing nominally. This of course meant a rescheduling of the LAM firings. The problem did not appear to be a serious one and had no implication for the satellite design itself.

COURTESY: ARIANESPACE
Prior to the launch, INSAT-2E at Arianespace's Preparation Building at Kourou.

The longer 75-minute LAMF-1 as per the original plan was in keeping with the satellite's higher lift-off mass of 2,550 kg as against 2,000 kg of the other INSAT-2 satellites. This would have been the longest ever firing in a nominal orbit-raising operation among the INSAT-2 satellites. This would have given a velocity increment of 59 per cent, with LAMF-2 and LAMF-3 planned for increments of 37 and 4 per cent respectively. Owing to a shorter firing duration, LAMF-1 has provided a velocity increment of only 12 per cent. LAMF-2 and 3 should have nominally taken place respectively on April 5 and 7. This sequence got altered and LAMF-2 was scheduled for April 6 when the satellite would be at its apogee for the eighth time instead of the nominal third apogee.

Under the altered plan, LAMF-3 may or may not be executed, according to Dr.S. Rangarajan, Director, MCF. If it is not executed a longer LAMF-2 will itself take the satellite into the desired orbit, and it will drift towards final location at 83° E. Since there is no unique sequencing of orbit-raising manoeuvres, the altered sequence will have no impact on the satellite's performance later in orbit, according to him. LASM firings will be followed by a few trimming manoeuvres using on-board thrusters, solar array and sail deployment, putting the satellite in the three-axis stabilised mode, switching on of the payload and their in-orbit tests. Following all these steps, the satellite is expected to be in its final slot after about two weeks of injection into the GTO by Ariane. If everything goes nominally, the satellite will be declared operational soon afterwards.

The launch window is fixed primarily on considerations of obtaining maximum electrical power from the onboard solar array and of getting sufficient earth sensor data for gyro calibration (necessary for proper attitude control). In a shared launch, there is always a compromise made on these counts because what is best for one satellite may not be the best for another. In a dedicated launch, these considerations can be optimised. Indeed, V117 did enable an earlier launch window for INSAT-2E as compared to earlier Ariane launches. Moreover, the injection was done without causing any tilt to the satellite, which allowed for better gyro calibration. Besides, owing to the greater injection velocity available (by about 19 metres per second ), a better transfer orbit was realised. The GTO parameters were 250 km x 36,155 km as against the usual 200 km x 35,936 km. More important, the orbit inclination was only 4° as against the nominal 7° in dual launches. The perigee advantage of 50 km and inclination advantage by 3° means an enormous saving of on-board fuel which otherwise would have been used for trimming manoeuvres.

T.L. PRABHAKAR
Inside the Spacecraft Control Centre at the Master Control Facility in Hassan, from where all operations of the satellite, including on-orbit operations, are monitored and controlled.

Besides, as Dr. Rangarajan pointed out, the fact that the transfer orbit apogee is greater than the geosynchronous value of 36,000 km lends flexibility in guiding the satellite into its slot during its drift phase. This also results in marginal fuel saving. The fuel thus saved will now be available for on-orbit corrections in the satellite's end-of-life phase, its nominal 12-year life will thus increase by about six months. Indeed, the design life is based on a value of the specific impulse (which is related to the rate at which fuel is consumed per unit thrust) for the LAM engine. Apparently, the performance of earlier satellites has shown that LAM has a better specific impulse, which means that less on-board fuel would have been consumed for the required velocity increases during LAMFs. From this perspective, the life-span of 2E may even be longer, up to 15 years.

Notwithstanding the fact that INSAT-2E will augment the telecommunications capacity provided by the INSAT series only in a limited way, it is perhaps the most advanced satellite built so far in terms of technologies used in the hardware and in the payloads. The communications payload has 17 C-band transponders, out of which five are in what is known as the 'lower extended C (LXC) band'. Ten of the channels, including the five in LXC, will have a shaped zonal beam whose coverage will include India, China, West Asia and a major part of South-East Asia. The remaining channels operate through a shaped wide beam whose coverage extends from Central Europe to Australia and includes China and the southern parts of Central Asia. With these, 2E will provide the largest communication coverage ever realised by a satellite in the INSAT series. Dual gridded reflectors, being used for the first time, enable the shaping of the beams in desired ways.

COURTESY: ARIANESPACE
The launch vehicle being rolled out during preparations for Flight V117.

The choice of LXC, as against UXC in the earlier INSAT-2 satellites, and the shapes of zonal and wide beams were apparently dictated by INTELSAT.

Given these footprints, uplinking to the spacecraft can be done from any point within the wide beam, not just from India. These transponders can provide 12,000 two-way voice links or broadcast about 3100 digitally compressed TV signals or a suitable mix of both.

COURTESY: ARIANESPACE
The launch complex ELA 3 in Kourou, from where the Ariane 5 launcher will carry INSAT-3B into orbit later this year.

According to R. Prasanna, regional representative, INTELSAT, all the leased transponders have been booked by television companies such as Eenadu TV, Sun TV and Vijay TV. Doordarshan has an interest in getting a larger share of the remaining eight C-band transponders on 2E in order to stay in competition with private channels and also have a reach beyond India. It may not be surprising if DoT and VSNL lease INTELSAT capacity for its operations, particularly for some value added services.

INSAT-2E is thus designed to continue and enhance the meteorological services which have been provided by INSAT-ID. It carries an improved version of the Very High Resolution Radiometer (VHRR) and a high resolution Charge Coupled Device (CCD) camera. Besides the usual visible and thermal IR band, the VHRR will have a new water vapour channel. In addition to imaging cloud cover (as is being done now) including the formation and movement of cyclones, the water vapour channel will aid the estimation of water vapour in the clouds and the forecasting of rain. As has been pointed out by Dr. T. N. Krishnamurti, the meteorological expert from Florida State University, the water vapour channel, an old tool in the U.S., European and Japanese satellites, would greatly enhance the study of monsoon. Also, for the first time over India, this will enable the estimation of upper tropospheric winds, those blowing 8 to 12 km above ground level.

The CCD camera is an offshoot of the successful experience of working with such cameras in the Indian Remote Sensing (IRS) satellites. This camera operates in three bands - visible, near IR and short wave IR - and with its higher resolution of 1 km, as against 2 km of VHRR, its data will complement the VHRR data and assist in cyclone analysis, local severe storm monitoring, heavy rainfall forecast and so on.

A short circuit in the power bus, suspected to have been caused by an anomalous charge build-up owing to external factors, led to a loss of earth lock in the use of INSAT-2D. This has led to several changes and contingency measures in the power system of 2E. Now, instead of a single power channel from the solar array, four streams of power channels have been incorporated so that in case of a short circuit only 25 per cent of the power bus would be lost. According to Dr. Rangarajan, any such loss of a part of the bus can at worst lead to a loss of seven transponders in the satellite. However, the bus has been so designed as to ensure that no critical subsystem is lost owing to the non-availability of power. Now it is possible to connect critical subsystems to any part of the bus by a ground command.

Unlike the silicon solar cells in the earlier satellites, 2E's solar array is made of high efficiency gallium arsenide/germanium (GaAs/Ge) based solar cells which have a solar power conversion efficiency of 18 per cent as against 12 to 13 per cent of silicon cells. This means a higher power to area ratio for the array and, coupled with the high-capacity (2 x 60 ampere-hour) nickel-hydrogen (Ni-H) batteries, it enables full payload support during eclipses as well. (Following an outage during an eclipse the stock exchange network in India had to be shut down recently. Earlier only 2 x 28 amp-hr nickel-cadmium (Ni-Cd) batteries were used.

Besides the CCD camera, the water vapour channel in the VHRR and the GaAs/Ge solar channels, many other products of state-of-the-art technology are being used in INSAT-2E for the first time. For example, in order to dissipate heat from hot spots rapidly, the northern and southern equipment panels of the satellite are embedded with thermal heat pipes filled with anhydrous ammonia. Efficient thermal management is important particularly with regard to Ni-H batteries whose performance is critically dependent on the temperature which has to be maintained at 10 °C. Application Specific Integrated Circuits (ASICs) and hybrid microcircuits (HMCs), designed in-house but fabricated in overseas silicon foundries, have been used for the first time and have helped reduce the volume and mass of the platform electronics packages, thus making the satellite bus more efficient.

The cuboid structure of INSAT-2E has been built around a corrugated thrust bearing central cylinder that has enabled the mounting of the propellant tanks - only one each for fuel and oxidiser - inside the cylinder itself for the first time. Earlier, in order to provide structural balance, a pair of tanks were used outside the central cylinder which consumed a lot of space. Varying radiation pressure on the solar array causes a season-to-season imbalance. Earlier this use to be countered by firing on-board thrusters. In INSAT-2E, a novel concept based on the principle of shape memory alloy has been employed in the design of two flaps at the end of the array that provides the counterforce.

Perhaps the most significant aspect of INSAT-2E is thus the fact that it is technologically a forerunner to the next series of satellites. INSAT-3 and INSAT-2E will mark the beginning of a new era of long-life mission satellites as ISRO enters a new millennium. This capability should also make ISRO a potential contractor in the world market for designing and building satellites. But in order to realise that potential, a commercial outlook, with quick turnaround times, is required. Maybe it is time to think of creating a corporate set-up under ISRO that functions as the prime contractor for satellites or even high-tech subsystems thereof.