ISRO’s Polar Satellite Launch Vehicle (PSLV) has long been regarded as one of the most reliable medium‑lift launchers in the world, but the recent back‑to‑back anomalies in its third stage have drawn attention to a structurally known weak point in multistage launch vehicles.

The failures on missions PSLV‑61 in May last year and PSLV‑62 last week have both been traced to malfunctions during the third of four stages, raising questions about what makes this phase particularly demanding, and whether any new risks have crept into an otherwise mature system.

The PSLV is a four‑stage rocket that alternates between solid and liquid propulsion. This architecture allows designers to exploit the strengths of each propellant type at different points in flight.

Solid motors, used in the first and third stages, offer high thrust and long storage life, making them ideal for the violent early climb and the powerful mid‑course acceleration needed to build orbital velocity.

Liquid stages, used in the second and fourth, provide precise control of thrust, better steering authority and the ability to fine‑tune trajectory, which is critical for accurate orbital insertion.

The first stage is a large solid booster responsible for lifting the vehicle off the pad and pushing it through the dense lower atmosphere. It burns for roughly two minutes, during which the rocket must tolerate intense aerodynamic loads, dynamic pressure and structural vibration.

Because the atmosphere is still relatively dense at this point, the vehicle is strongly stabilised aerodynamically, and the guidance, navigation and control system can rely on both thrust vectoring and aerodynamic surfaces to maintain attitude within tight margins.

Once the first stage burns out and separates, the second stage, a liquid‑fuelled module, takes over. Its engine provides smoother, more controllable thrust and allows finer adjustments in pitch, yaw and roll. 

With atmospheric drag and dynamic pressure already significantly reduced, the heat shield can be safely jettisoned, shedding dead mass and improving performance. The second stage typically operates for a few minutes, continuing to raise the rocket’s altitude while shaping the flight path to prepare for the more orbit‑centric work of the upper stages.

The third stage marks a return to solid propulsion, but under conditions very different from those encountered by the first stage. Ignited at high altitude or near vacuum, it must provide a strong, relatively short burn that transitions the vehicle from a predominantly vertical climb to a high horizontal velocity, which is the key requirement for entering orbit.

Unlike the first stage, it no longer enjoys the stabilising influence of a thick atmosphere, so the vehicle’s attitude is governed almost entirely by its internal control systems, such as thrust vector control, reaction control thrusters and sophisticated guidance algorithms.

Because the third stage uses a solid motor, it lacks the fine throttle and shut‑off capability of liquid engines. Once ignited, a solid motor’s thrust profile is largely predetermined by grain geometry and propellant characteristics, meaning that any anomaly in chamber pressure or thrust cannot be corrected by simply throttling back or restarting.

If pressure drops or thrust becomes unsteady, the guidance system has limited tools to compensate, particularly in near‑vacuum conditions where aerodynamic control surfaces are ineffective. This rigidity in operation makes solid upper stages intrinsically less forgiving than their liquid counterparts.

Compounding this is the fact that the third stage operates at a crucial transition point in the mission. By this time, much of the rocket’s mass has been discarded, so the upper composite—the remaining stages plus payload—has a much higher mass‑to‑torque sensitivity.

Even small perturbations in thrust direction or rotational motion can lead to more pronounced attitude deviations. If roll, pitch or yaw rates begin to grow faster than the control system can counter, the vehicle may experience a loss of attitude control, potentially compromising the subsequent ignition of the fourth stage and the accuracy of orbital injection.

The PSLV‑61 anomaly illustrates these sensitivities. Although ISRO has not publicly released the full fault analysis, it is known that the vehicle suffered a drop in pressure in the third‑stage propulsion chamber.

This drop meant that the motor could not maintain the required thrust, leading the rocket to deviate from its planned trajectory. Without the expected energy contribution from Stage 3, the vehicle’s guidance solution and timing for later events, such as fourth‑stage ignition and satellite separation, were rendered ineffective, resulting in mission failure.

In the case of PSLV‑62, the malfunction unfolded later in the third‑stage phase. According to statements from ISRO officials, including Chairman V. Narayanan, the rocket performed nominally through the first two stages and almost the entire third stage.

The anomaly emerged as disturbances in the vehicle’s attitude, leading to excessive rolling of the upper composite. Former ISRO scientist Manish Purohit described this as a “loss of attitude control”, where the rolling rate of the upper stages increased beyond manageable limits just as the fourth stage was preparing to ignite.

When the fourth stage eventually fired and separated, it did so while the upper stack was no longer pointing in the correct direction. Although the fourth stage is a liquid‑fuelled module with good steering authority, it can only correct deviations within a certain envelope.

If the initial orientation at ignition is significantly off‑axis, the stage ends up expending its limited propellant merely trying to stabilise and redirect itself, leaving insufficient capability to attain the planned orbit. In PSLV‑62’s case, this resulted in the effective loss of 15 of the 16 satellites on board, with the lone surviving satellite only briefly transmitting before contact was lost.

From a systems perspective, the problem is not that the third stage is inherently unreliable, but that it occupies a high‑leverage position in the mission profile with comparatively fewer corrective options. Its solid nature reduces operational flexibility, its high‑altitude environment strips away aerodynamic stability, and its timing makes it critical for setting up the fourth stage. Any shortfall or instability at this point propagates directly into the final orbital insertion phase, often with no opportunity for recovery.

Another complicating factor is that, although PSLVs share a common design, each flight uses hardware that is essentially one‑off. The vehicle is expendable and is built from components manufactured, assembled and integrated afresh for each mission. Even in a stable configuration with no major design changes, minor variations in materials, fabrication tolerances, integration processes or supplier chains can lead to subtle differences in performance. Over dozens of flights, such variations normally sit well within margins, but occasionally they can interact unfavourably with environmental factors or operational conditions to produce rare anomalies.

ISRO has convened fault assessment committees to investigate both recent failures. The findings of the PSLV‑61 probe have not been made public, and no preliminary details about PSLV‑62 have been released beyond the acknowledgement of attitude control loss during the third stage.

Analysts such as Manish Purohit have noted that it is not yet clear whether any corrective measures proposed after the PSLV‑61 anomaly were implemented, or whether they relate in any way to what transpired on PSLV‑62. Until ISRO publishes at least a high‑level summary, it is difficult for external experts to assess whether these incidents share a common root cause or are unrelated aberrations.

Historically, PSLV’s record argues strongly for the latter interpretation. Prior to missions 61 and 62, PSLV had completed over 60 flights with only a handful of partial or total failures, giving it one of the best reliability statistics in its class worldwide.

It has delivered landmark missions such as Chandrayaan‑1 in 2008, which propelled India’s first spacecraft to the Moon, and Mangalyaan in 2013, which made India the first Asian nation to reach Mars orbit and the first in the world to do so on a maiden attempt. In 2017, PSLV‑C37 set a world record by deploying 104 satellites in a single launch, demonstrating not only reliability but also complex mission management capabilities.

Over the decades, PSLV has also become a preferred launcher for small and medium satellites from international customers, a status that would not have been possible without consistent performance. This legacy underscores why the recent failures are being treated as anomalies rather than indicative of systemic decline.

For a rocket with such a long and successful operational history, two failures, even in quick succession and on the same stage, warrant serious technical scrutiny but not conclusions about an enduring design flaw without evidence.

Nevertheless, the clustering of anomalies in the third stage suggests that this phase will attract heightened engineering attention. Potential areas of focus include detailed reassessment of solid motor grain design and manufacturing controls, enhanced non‑destructive evaluation of propellant and casing, refinement of thrust vector control hardware and software, and improvements in attitude determination and control margins during low‑aerodynamic‑pressure flight. Engineers may also explore more robust fault detection and isolation algorithms that can rapidly identify off‑nominal behaviour during Stage 3 and trigger contingency modes for the fourth stage.

Given the growing complexity of missions, including multiple satellite deployments, rideshare configurations and more demanding orbits, the pressure on upper‑stage performance continues to increase.

This makes robust upper‑stage design and control, especially in solid‑propelled segments like PSLV’s third stage, a central priority. Any incremental improvements derived from the ongoing investigations will not only restore confidence in PSLV but also feed forward into newer ISRO launchers such as SSLV and the Gaganyaan‑class human‑rated vehicles, where reliability requirements are even more stringent.

For now, the third stage remains a technically challenging but well‑understood element of the rocket, one whose difficulties stem from fundamental characteristics of solid propulsion and high‑altitude dynamics rather than negligence or poor design.

ISRO’s task is to identify whether recent events point to a specific correctable weakness—be it in manufacturing, quality assurance, control systems or operational margins—and to implement targeted fixes without overcorrecting a system that has served India’s space ambitions remarkably well.

IDN (With Agency Inputs)