Value Stream Mapping for Industrial Gas Cylinder Filling and Distribution: From Order Release to Dock-Ready Cylinder Without the Asset Churn

1. The Industrial Gas Value Stream: Why the Cylinder Float Matters

In the realm of industrial gas distribution, value is created when the customer receives a correctly filled, correctly labelled, safely compliant cylinder or bulk delivery on the agreed dock date.

The physical filling operation is visible, measurable and often treated as the primary capacity constraint. However, many plants have a different limiting factor: the availability and location of usable cylinder assets.

A cylinder may be physically present but unavailable because it is:

• Awaiting return from a customer
• Mis-sorted into the wrong gas family
• Due for statutory inspection or testing
• Lost within a depot or remote storage lane
• Awaiting manual reconciliation
• Held in a finished-goods location without a confirmed demand signal

This is asset churn: repeated searching, handling, relabelling, sorting and relocating of cylinders without increasing customer value.

Value Stream Mapping (VSM) makes that hidden flow visible. As outlined in the Lean Six Sigma concepts and glossary, the purpose is to view material and information flow as one connected system rather than optimise isolated process steps.

2. Scope Selection: Define the Family Before Drawing the Map

This guide maps the industrial gas cylinder family from customer order release to dock-ready delivery and empty-cylinder return.

The selected family is:

Oxygen, argon, nitrogen and CO2 cylinders operating on a 41,000-cylinder float.

The scope begins with:

1. Customer order entry and cylinder asset retrieval
2. Empty-cylinder return and receipt
3. Test-due sorting and inspection status verification
4. Evacuation and purge
5. Filling at the assigned manifold
6. Weight, pressure and gas analysis checks
7. Valve and label verification
8. Palletisation and staging
9. Secondary distribution to the customer dock
10. Empty-cylinder return to the plant

The scope excludes medical gas validation, clinical traceability requirements and on-site bulk plant engineering. Those are separate value streams with different regulatory, validation and process-control requirements.

The mapping boundary should be approved before data collection. A defined boundary prevents the team from expanding into every warehouse, transport and engineering activity. The Project Scope Boundary Calculator can support this Define-phase decision.

3. Current-State Mapping: Where Flow Breaks Down

A practical current-state map should contain process boxes, inventory triangles, information arrows and a bottom timeline separating value-added work from waiting.

For this example, the process boxes are:

Information flow typically begins with order entry in the ERP or customer service system. It then moves through a planning schedule, a partially updated cylinder tracking system, handwritten or printed stickers, and manual reconciliation at dispatch.

The physical flow is less direct. Cylinders move between return lanes, test-due areas, gas-specific staging, fill bays, quality-control locations, palletisation and transport staging. Every handoff creates an opportunity for misidentification, waiting or unnecessary motion.

4. Worked Example: Trace 1,000 Cylinders Through One Week

Assume the map follows 1,000 cylinders through a representative week.

Key observations include:

• Fill manifold cycle: 38 seconds per cylinder
• Manifold uptime: 87.6%
• Fill rework: 4.1%, or 41 cylinders
• Mis-sort rate: 6.3%, or 63 cylinders
• Cylinder-finding time during a bay change: 22 minutes
• Test-due backlog: 9.8%, or 98 cylinders
• Cylinder touches: 1,850 per 1,000 filled
• Order-to-delivery lead time: 4.6 days
• Float utilisation: 71%

Total elapsed lead time

Convert the order-to-delivery lead time into hours:

4.6 days × 24 hours/day = 110.4 hours

Or, in minutes:

4.6 × 24 × 60 = 6,624 minutes

Value-added fill time

The nominal filling time is:

1,000 cylinders × 38 seconds = 38,000 seconds

38,000 ÷ 3,600 = 10.56 hours

Therefore, the direct fill time is 10.56 hours, or approximately 633 minutes.

The 41 reworked cylinders add:

41 × 38 seconds = 1,558 seconds = 0.43 hours

That 0.43 hours is rework and should not be counted as first-time value-added work.

Process Cycle Efficiency

Using direct successful fill time as the value-added numerator:

PCE = Value-added time ÷ Total lead time × 100

PCE = 10.56 ÷ 110.4 × 100 = 9.57%

The process is therefore spending approximately 90.43% of elapsed time in waiting, movement, queues, scheduling, verification or other non-value-added activity.

This is not a criticism of the operators. It is a system signal. In the Analyse Phase of DMAIC, the team should use average queue time, attribute data such as Pass/Fail results, and visual tools such as Pareto charts and box plots to separate recurring causes from isolated events.

5. The Eight DOWNTIME Wastes in the Fill Plant

Using an illustrative annual volume of 52,000 cylinders-1,000 cylinders per week for 52 weeks: the following exposure becomes visible. Local labour and asset rates should replace these planning estimates.

• Defects: At 4.1% rework, approximately 2,132 cylinders per year require additional work. At an estimated $18 per rework event, that represents approximately $38,000.
• Overproduction: Filling 300 cylinders ahead of confirmed demand each week creates 15,600 cylinder-turns per year of premature inventory and handling.
• Waiting: A 9.8% test-due backlog represents roughly 98 cylinders in the weekly example waiting for release, reducing delivery responsiveness.
• Non-utilised talent: If coordinators spend 1.5 hours per day searching, reconciling and correcting asset records, 750 hours annually may be redirected from planning and improvement work, approximately $31,500 at a $42 loaded hourly rate.
• Transportation: Moving cylinders between remote storage, incorrect bays and dispatch lanes can consume approximately 1,200 avoidable cylinder-turns per year.
• Inventory: At 71% float utilisation, 11,890 of the 41,000 cylinder positions are not actively supporting productive turns at a given point in time.
• Motion: Three bay changes per day at 22 minutes each creates approximately 275 hours annually of finding and repositioning activity, or about $11,600 in labour.
• Extra-processing: Sticker replacement, duplicate checks and manual reconciliation can consume two coordinator-hours per dispatch day, approximately $17,500 annually.

6. Future-State Build: Pull the Cylinder Float

The future state treats the cylinder asset as a controlled kanban loop, not as anonymous inventory.

The design should include:

• Barcode tracking at return, test release, fill completion and dispatch
• Dedicated fill lanes by gas type and test status
• 5S locations with clear visual labels and ownership rules
• Pull-based replenishment of the 41,000-cylinder float based on customer consumption
• FIFO lanes for returned empties
• SMED methods for manifold bay changes, including pre-staged connectors, tools and labels
• Standard work at the manifold for connection, fill, verification and release
• In-line weight and gas analysis checks at the point of fill
• A live dock board showing order, cylinder status, route, carrier and readiness

The operating logic is straightforward: customer consumption creates the signal; the signal pulls a cylinder through the loop; each scan confirms its status and location.

The future state should also be checked against takt time:

Takt time = Available production time ÷ Customer demand

Throughput should be increased by improving the constraint, not by filling cylinders that are not required. If the constraint is test release or asset retrieval, adding theoretical manifold capacity will not improve end-to-end delivery.

7. Current State vs Future State

The following future-state figures are practical improvement targets for the worked example, not audited plant results.

A live X-bar chart for average fill or release time, supported by an R chart for short-term range, can help confirm whether improvement is stable or merely temporary. Variation should guide action: common-cause variation calls for system redesign, while special-cause variation requires targeted investigation.

8. 90-Day Kaizen Sequencing

Days 1–30: Stabilise and see the flow

Actions

• Confirm the family, boundary and customer demand profile.
• Time-stamp 1,000 cylinders from return to dispatch.
• Establish FIFO lanes for empties and test-due cylinders.
• Audit the 6.3% mis-sort category.
• Create a visual daily board for backlog, rework and dock readiness.

Owners

• Plant manager: scope, safety and governance
• Fill-line supervisor: standard work and time studies
• Distribution coordinator: order-to-dock data

Metric expected to move: mis-sort rate, test-due backlog and data completeness.

Days 31–60: Create pull and reduce changeover loss

Actions

• Introduce barcode scans at the four critical status changes.
• Pilot a kanban loop for one gas family.
• Run a SMED workshop on the highest-frequency bay change.
• Standardise manifold connection, weight and analysis checks.
• Use a daily average and Pareto of rework causes.

Owners

• Plant manager: pilot approval and resource removal
• Fill-line supervisor: SMED and standard work
• Distribution coordinator: consumption signal and route sequence

Metric expected to move: cylinder touches, changeover time, rework and float utilisation.

Days 61–90: Lock in the future state

Actions

• Expand barcode tracking across all four gas families.
• Link the dock board to order and cylinder status.
• Review the future-state map against takt and throughput.
• Confirm control-plan ownership and escalation rules.
• Recalculate PCE and validate the cost-per-cylinder target.

Owners

• Plant manager: control plan and benefits review
• Fill-line supervisor: process adherence and control charts
• Distribution coordinator: customer delivery performance

Metric expected to move: lead time, PCE, on-time delivery and cost per cylinder filled.

9. Build the Capability to Improve the Value Stream

This example demonstrates how current-state mapping, future-state mapping, kanban, SMED and Process Cycle Efficiency convert operational complexity into a measurable improvement plan.

For professionals responsible for manufacturing, logistics, quality or industrial distribution, these skills provide a practical route from asset churn to controlled flow. Lean Six Sigma training also connects VSM to DMAIC, root-cause analysis, control plans, capability measures and financial justification.

Build these skills through CSSC-accredited Lean Six Sigma Green Belt or Black Belt training at lean6sigmahub.com. Learn how to map the current state, design the future state, quantify PCE, build kanban systems and lead measurable improvement projects.

Kaizen. Kai-Care. Kai-Done. Lean Six Sigma