Industrial Safety Is Not a Product. It Is a System of Work
Industrial Safety Is Not a Product. It Is a System of Work.
Three Fatalities at Sandhar Technologies, Gurugram — A Larger Question
for Industrial Management and Safety Professionals
HSE-RM SOLUTIONS | Professional Perspective | 13 November 2026
Three workers have lost their lives in an incident involving a tank at an
automotive components facility in Gurugram. The investigation must establish
what happened and who was responsible. For other industries, the immediate
lesson is to examine how hazardous work is planned, authorised and controlled.
Adequate safety arrangements mean more than equipment. They mean a
system that prevents people from facing hazards without effective protection.
The incident and what is currently known
News reports describe three worker deaths following an incident on 11
November 2026 at Sandhar Technologies in Dhumaspur, near Bhondsi, Gurugram,
Haryana. The workers became unconscious in a tank described in reporting as a
chemical tank. Other employees removed them and took them for medical
treatment.
The Times of India reported maintenance-related entry and a complaint
alleging inadequate safety arrangements. Bhondsi police registered an FIR
against company management under Sections 106 and 3(5) of the Bharatiya Nyaya
Sanhita. These are reported allegations under investigation. Registration of an
FIR is not a finding of guilt.
The entry sequence remains uncertain. An earlier report described
workers falling into the tank, while later reporting described workers being
directed to enter. The exact task, tank contents, hazardous atmosphere and
sequence must therefore be established from evidence.
No final technical investigation report was located in the sources
reviewed through 13 November 2026. Those reports also do not confirm arrests.
The discussion below sets out questions and safeguards for comparable work; it
does not establish which controls were absent or ineffective at this facility.
Three families have suffered an irreversible loss. They deserve a
thorough investigation and accountability based on evidence. Industry also has
a responsibility to learn before another worker faces a similar danger.
1 Safety cannot be purchased as a complete solution
Equipment, instruments and technology can be essential to safety. The
problem begins when purchasing them is treated as proof that the risk has been
controlled.
A detector must suit the substances that could be present. A permit must
reflect the actual task and conditions. A rescue system must match the space,
the emergency and the competence of the people expected to use it. A dashboard
must contain reliable information that leads to action.
Each product belongs within a larger system of work. That system must
define what can go wrong, which controls are necessary, who is responsible and
how the controls will remain effective.
What are we trying to prevent, and what must work reliably to prevent
it?
That question should come before a purchase order. Once the risk is
understood, equipment can be selected for a clear purpose and its performance
can be checked.
2 Practical experience and modern engineering belong together
Experienced operators and maintenance personnel often know details that
are missing from a procedure. They may know which line tends to block, which
valve has passed in the past, where residue collects or how equipment behaves
after a shutdown.
This knowledge should be captured and examined. It can improve the risk
assessment, reveal differences between drawings and the actual plant, and help
designers make controls easier to use.
Experience also has limits. A task may have been completed many times
without an incident while a serious hazard remained present. Familiarity can
make an unsafe condition seem normal.
Earlier generations of industry had valuable practical knowledge, but
they also experienced serious accidents. Modern engineering, detection and
protective systems provide benefits that should be retained. The useful
approach is to combine operating experience with sound engineering, verified
information and disciplined work methods.
Workers should be encouraged to ask why a condition has changed.
Management should make sure those questions receive a competent response.
3 Understand the task before selecting the controls
A request to “clean the tank” or “carry out maintenance” is too vague to
support a sound decision. The first step is to define the task, the equipment
involved and the conditions under which the work will take place.
Before authorising work, establish:
- What the tank contains
now and what it contained previously.
- Which residues,
cleaning agents or reaction products could be present.
- Whether a hazardous
atmosphere could exist or develop.
- Which pipes, pumps,
agitators and energy sources connect to the tank.
- Whether the work can
be completed without entry or exposure at the opening.
- What may change when
the tank is opened, drained, cleaned or disturbed.
- Who will assess,
authorise, perform and supervise the task.
- How people will
withdraw and how rescue will be provided if needed.
The name of the industry does not identify the tank hazard. An
automotive components factory may have several different processes and
utilities. The specific tank service, chemicals and maintenance method must be
established. A generic label such as “chemical tank” is not enough to select a
detector, respirator or cleaning method.
Recognise the full range of hazards
Atmospheric hazards deserve close attention, but the assessment must also
consider unexpected inflow, chemical contact, moving equipment, stored energy,
heat, falls and difficult access where relevant. Exposure can occur at an
opening as well as during entry.
Assess a credible event, rather than writing only “confined-space risk.”
For example: a person approaches an open tank, encounters a hazardous
atmosphere and collapses. Another scenario may involve liquid entering through
a connected line during internal work. Each scenario needs controls that
address its own cause and consequences.
Give potential fatalities the attention they require
Consider severity, likelihood and exposure using the organisation’s
approved assessment method. An infrequent task can still have fatal
consequences. An unsupported assumption that failure is unlikely must not
become permission to proceed with an uncontrolled hazard.
Missing information about a critical condition is a reason to resolve
that uncertainty before authorisation. A proposed control should not reduce the
recorded current risk until it has been implemented and verified.
Apply the hierarchy of controls
The preferred order is elimination, substitution, engineering
controls, administrative controls and personal protective equipment. For
tank maintenance, consider the following applications.
|
Control level |
Application to consider |
|
Elimination |
Complete the task externally or redesign the arrangement so that
entry is unnecessary. Assess hazards at the opening as well. |
|
Substitution |
Where technically feasible, use a less hazardous material or
cleaning method. Check compatibility and any new risks. |
|
Engineering controls |
Provide suitable isolation, safe access, ventilation and other
measures that control the identified hazards. |
|
Administrative controls |
Define competent authorisation, permits, supervision,
communication, monitoring and stop-work conditions. |
|
PPE |
Select protection for the actual exposure and task. Confirm
suitability, fit, training and maintenance. |
Process isolation needs its own plan. Identify connected flows and stored
energy, select an appropriate isolation arrangement and verify it. Stopping a
pump or closing a single valve does not automatically establish safe isolation.
Rescue planning supports these controls; it is not another level in
the hierarchy. Before exposure, establish a rescue arrangement suited to
the particular space and foreseeable emergency. Consider non-entry retrieval
where suitable and ensure trained, equipped responders can act within the time
the hazard demands. Unprotected rescue entry can expose another person.
Where an essential safeguard is unavailable or cannot be verified,
postpone the work. Budget and schedule constraints do not make an unsafe
condition acceptable.
4 Observe the work and monitor the conditions that matter
A practical system must connect what people see, what instruments measure
and what the team does next. HSE-RM SOLUTIONS describes this daily discipline
as Observe, Monitor and Control.
Observe
Examine the actual workplace and task. Look at the tank openings,
connected lines, access arrangements, nearby activities and the way people are
preparing to work. Compare the physical arrangement with drawings and
procedures.
Speak with the people doing the job. Ask them to explain the work
sequence, the hazards and the conditions that would make them stop. If the
explanation differs from the permit, resolve the difference before work begins.
Monitor
Select measurements from the hazard assessment. For a tank task, these
may include oxygen concentration, flammable atmosphere readings, relevant toxic
substances and ventilation performance. Isolation condition, connected
operations and entry records also need appropriate checks.
Atmospheric testing must use equipment capable of detecting the
contaminants that could be present. Confirm the instrument’s condition,
calibration and required pre-use checks. Define the sampling locations, timing,
instrument response time and acceptable conditions. Record what was measured,
the results, the time, location and person responsible. “Gas test completed” is
not a sufficient account of the condition.
A normal oxygen reading does not establish that the air is free of toxic
substances. Likewise, a standard multi-gas detector does not detect every
chemical. Competent assessment must determine what additional testing is
needed.
Ventilation must use suitable clean air and avoid creating another
exposure at its discharge. Never use oxygen to ventilate a confined space.
Ordinary masks and air-purifying respirators do not supply oxygen. Respiratory
protection must be selected for the atmosphere and used within a competent
respiratory protection programme.
Control
The work team must know what each result requires them to do. The
following are examples to build into a site-specific procedure:
|
Condition |
Required
decision |
|
Tank contents
or atmospheric hazards remain unresolved |
Do not
authorise entry until the assessment establishes the necessary controls. |
|
Isolation
cannot be verified |
Keep the work
stopped and resolve the isolation. |
|
An alarm
occurs, ventilation fails or conditions become unsafe |
Withdraw as
required by the procedure and reassess before re-entry. |
|
Communication
or required rescue readiness is lost |
Suspend the
entry and restore the required arrangements before reauthorisation. |
|
The task or
connected operation changes |
Review the
assessment and permit before continuing under changed conditions. |
After corrective action, verify the result. At completion, account for
people, close the permit and reinstate equipment through a controlled handover.
Review recurring problems so that the next job benefits from what was learned.
This daily approach belongs within a wider cycle of planning,
implementation, checking and review. Observe, Monitor and Control cannot
replace that full management cycle.
5 Collect information that supports decisions
A facility may generate hundreds of permits, inspection forms and
training records. Their value depends on the decisions they support.
For example, a monthly report showing “50 confined-space permits issued”
describes activity. It does not show whether the tasks were safe.
More useful questions include:
- How many tasks avoided
entry after the work method was reviewed?
- How many needed
additional isolation or other safeguards?
- Which tasks stopped
because conditions were unsafe or uncertain?
- What changed during
work, and how did the team respond?
- Which defects or
deviations appeared repeatedly?
- Were corrective
actions completed and their effectiveness verified?
Interpret the results carefully. More reported stoppages may reflect
better recognition of hazards and stronger use of stop-work authority. A target
of zero stoppages could discourage the very decisions the system needs.
Preventive information should be considered alongside incident data.
An example of using operational information
Consider a facility that repeatedly enters a process tank for cleaning.
This is an illustrative example, not a finding about Sandhar Technologies.
The team records why entry is needed, where residue collects, the time
spent cleaning, relevant atmospheric results and interruptions. A review may
identify a suitable external cleaning arrangement or a design change that
reduces the need for entry.
The proposed method must still be assessed. External cleaning can
introduce hazards such as pressure, chemical exposure, moving equipment or
unexpected discharge. Any improvement should demonstrate a reduction in the
relevant risk without leaving new hazards uncontrolled.
The useful outcome is a justified decision about the work method. A
larger spreadsheet or a more attractive dashboard is valuable only if it helps
the team reach and implement that decision.
6 Work within constraints without weakening essential safeguards
Safety professionals need to understand the organisation’s operating and
financial limits. They also need to identify which conditions must be met
before work can proceed.
A useful recommendation explains:
- The hazard and
credible consequence.
- The existing controls
and evidence of their condition.
- The critical gaps and
immediate operating restrictions.
- The options available
to reduce the risk.
- The resources,
competence and maintenance each option needs.
- The person
responsible, completion date and verification method.
This gives management a basis for deciding where to invest. Some
improvements may involve changing the work method, improving scheduling,
clarifying responsibility or completing overdue maintenance. Others require
engineered modifications, suitable instruments, ventilation or specialist
rescue capability.
Compare options that can meet the required safety standard. Include
installation, training, maintenance, testing and replacement costs, as well as
the purchase price. An apparently inexpensive option may be difficult to
maintain or unreliable in actual use.
Cost optimisation means removing ineffective expenditure while
providing the safeguards the work requires.
Where a permanent improvement takes time, competent people must decide
whether a valid temporary arrangement can control the risk. If it cannot,
restrict or stop the activity. A completion date in an action register does not
protect someone doing the job today.
7 Select a solution for the actual facility
A product’s use at another large company can be useful background
information. It does not establish suitability for every application.
Two factories making similar products may have different chemicals, tank
arrangements, staffing, maintenance practices and exposure conditions. Their
teams may also have different capacity to inspect, calibrate, repair and use
the equipment.
Before adopting a system, ask:
- Which specific
scenario will it prevent or control?
- Is that scenario
relevant to our actual operation?
- Will the system work
across our expected operating conditions?
- What training,
maintenance and specialist support will it require?
- What happens if it
loses power, fails or gives an uncertain result?
- How will we
demonstrate that it performs as intended?
- Is there a simpler
compliant option with equivalent or better protection?
A system should be understandable to the people who use it. Instructions
should suit their language and literacy, and responsibilities should be clear
during shift changes and contractor work.
Simpler administration can improve use. It must not remove a necessary
test, isolation, authorisation or emergency arrangement. The aim is to make
essential controls easier to apply correctly.
8 Management responsibility extends beyond procurement
Management must provide the conditions in which safe work can happen.
That includes suitable engineering, competent people, maintenance resources,
realistic schedules and authority to stop a task.
A signed permit or completed training register is useful evidence, but
investigators must also establish what happened at the workplace.
Questions for a fair investigation
|
Investigation
area |
Questions to
establish from evidence |
|
Task and
timeline |
What work was
planned, who assigned it and what happened before the first loss of
consciousness? Did people enter, fall or approach an opening? |
|
Process and
atmosphere |
What did the
tank contain, what had recently changed and what could explain the exposure?
What do samples, operating records and medical findings support? |
|
Safeguards |
Which
controls were required, which were present and what evidence shows their condition
and use? |
|
Work
organisation |
Were roles
clear, did the team have suitable resources and supervision, and could
workers stop or question the task? |
|
Emergency
response |
How was the
alarm raised, how did recovery take place and were additional people exposed? |
|
Earlier
learning |
Were similar
tasks, defects, warnings or near misses previously recorded? What action
followed? |
Preserve evidence through the authorised investigation team once safe
access is established. Compare witness accounts with permits, work orders,
instrument records, isolation records, CCTV and process information where
available. Readings taken after ventilation, draining or rescue may not
represent the original conditions.
Examine alternative explanations. Do not infer that welding occurred
merely because a worker’s trade was welder, or identify a gas solely because it
is associated with some other tank incident.
Accountability and learning support each other when both follow evidence.
An early label such as “human error” can leave important questions unanswered.
The investigation must examine the decisions and conditions that made the
action possible, while establishing individual and organisational
responsibilities fairly.
Safety culture appears in everyday decisions
Culture is shaped by management choices, industry practices and the
conditions in which people work. It includes whether defects receive attention,
whether employees can raise concerns without fear, and whether a production
deadline overrides a safety limit.
Workers have an important role in identifying problems, but they cannot
be expected to compensate for unsuitable equipment, missing engineering
controls or an unworkable plan. Safety professionals support this system;
operational managers remain responsible for controlling the work within their
roles.
9 How HSE-RM SOLUTIONS supports practical system development
At HSE-RM SOLUTIONS, we begin by understanding the operation and the
risks it creates. We then help the organisation develop controls and work
arrangements that its teams can use, maintain and verify.
Confined-space risk assessment
We review the task, potential atmospheres, connected systems, access and
exposure routes. The assessment considers alternatives to entry, isolation
needs, foreseeable changes and the conditions required before authorisation.
Maintenance and permit-to-work system review
We examine how work moves from planning to completion. This includes
coordination between operations, maintenance, contractors and supervisors, as
well as isolation, permit checks, monitoring, suspension and handover.
Emergency rescue preparedness and training
We assess foreseeable rescue scenarios, equipment suitability, responder
competence and coordination. The aim is to establish practical readiness for
the spaces and tasks involved, supported by appropriate training and exercises.
Operational information and safety system development
We help teams decide what information to collect, how to check its
reliability and how to use it. Inspection and reporting should support clear
decisions, assigned actions and verified completion.
Our BRMS and DRS service approach
Through our BRMS and DRS service models, professionals study day-to-day
activities and collect relevant operational information. Technical and
analytical teams then review that information to identify hazards, recurring
deviations and opportunities for improvement.
The approach connects site observations with management decisions:
- Understand the work
and identify the information needed.
- Collect and validate
observations, measurements and operating records.
- Assess the risks and
identify weaknesses in controls.
- Recommend priorities,
responsibilities and practical improvements.
- Help the organisation
establish follow-up and verification arrangements.
The intended output is a clear account of where serious risks exist,
which actions need priority, and how the organisation can check progress.
Recommendations should explain their safety purpose and resource needs before
management commits to a solution.
10 Questions to ask before the next tank job
The Gurugram investigation must establish the actual cause and the
adequacy of the arrangements at the site. Other organisations can review
comparable tasks immediately without assuming that the same mechanism applies
to them.
Before the next job, management and the work team should be able to
answer:
- Do we understand the
tank’s contents, connections and credible hazards?
- Have we assessed a
method that avoids entry or reduces exposure?
- Which conditions must
be established and verified before work starts?
- Who will monitor those
conditions, and what results require action?
- Can the team explain
the stop-work and withdrawal criteria?
- Is a suitable rescue
arrangement ready for this task and shift?
- Who can authorise
restart after a change or interruption?
If a critical answer is missing, the next step is to resolve it before
exposing people to the hazard.
These questions help turn a general commitment to safety into a decision
that can be checked at the workplace.
Final perspective
Industrial safety depends on understanding the operation, anticipating
credible failures and keeping effective controls in place. Equipment and
documents support that work when their purpose is clear and their performance
is verified.
A sound system gives people the information, resources and authority they
need to act safely. It is understandable, technically reliable and maintainable
throughout its use.
The objective is to make safety effective. Where the possible
consequence is fatal, essential safeguards are not optional.
At HSE-RM SOLUTIONS, we help organisations build practical systems that
connect hazard identification with daily operating decisions. The value of a
safety investment lies in the protection it provides and its ability to keep
providing that protection.
HSE-RM SOLUTIONS
Your Partner In Safety...

Comments
Post a Comment