In potable water, value is not created by compliance alone. Customers experience value as safe, available, consistent water: and confidence that the supply will remain reliable when conditions change.
That makes a water treatment plant an ideal environment for value stream mapping. The method reveals how raw water, information, chemicals, equipment capacity and decisions flow from abstraction to the customer. It also exposes where waiting, variation, rework and losses weaken the service before a customer ever notices them.
A useful value stream map connects operational performance with customer confidence:
- Is treated water consistently below the 0.3 NTU turbidity target?
- Is residual chlorine stable across the network?
- How quickly can the plant respond to raw-water changes?
- How much treated water is lost through leakage or metering gaps?
- How often do valve faults interrupt flow?
- Do operators receive the right information at the right time?
The EPA Lean Water Toolkit and established Lean guidance on value stream mapping provide a sound foundation. The following deep guide applies those principles to a potable water treatment operation.
1. Define the water value stream and its boundaries
The proposed scope begins with:
- Raw water abstraction and intake pumping
- Screening
- Coagulation and flocculation
- Sedimentation
- Filtration
- Disinfection
- Reservoir storage
- Distribution monitoring
The boundary must be explicit. A plant-level map may end at the treated-water outlet, while an extended customer-level map may continue through reservoirs, trunk mains, district-metered areas and customer meters.
For a plant serving a remote site or campground, the boundary may reasonably extend to the campground reservoir, pressure zone and final monitoring point. For a municipal plant, the map may stop at the plant outlet, with non-revenue water analysed in a connected distribution map.
This prevents a common failure: attempting to map every asset at once and producing a diagram too broad to support action. Use a macro map to show the complete stream, then create micro-maps for bottlenecks such as filtration, chemical dosing or distribution leakage.
2. Calculate takt time for a continuous process
The site in this worked example supplies 45 ML per day. With 1,440 available plant minutes per day:
[
\text{Takt time}=\frac{1,440\text{ minutes}}{45\text{ ML}}=32\text{ minutes per ML}
]
In a continuous process, takt time is not a batch instruction. It is a pacing reference: the plant must produce approximately 1 ML every 32 minutes to match average demand.
The map should distinguish between:
- Cycle time: the effective processing time associated with each stage
- Flow time: the elapsed time for water to move through the system
- Lead time: total elapsed time, including waiting, storage and release decisions
- Available capacity: what the process can deliver during the demand window
3. Build the current-state value stream map
A cross-functional team should walk the process and capture actual conditions rather than relying only on standard operating procedures. Include operations, laboratory and quality staff, maintenance, SCADA or automation specialists, distribution personnel and regulatory reporting owners.
The information flow is as important as the water flow. Map:
- SCADA signals for flow, turbidity, pressure, level and chlorine residual
- Laboratory sampling schedules and result-release times
- Regulatory reporting requirements and approval points
- Chemical ordering and inventory signals
- Preventive maintenance schedules
- Corrective maintenance work orders
- Operator escalations and alarm responses
Worked current-state example
The following figures are illustrative but realistic for a 45 ML/day plant:
| Process stage | Effective process time | Key operating condition |
|---|---|---|
| Abstraction and screening | 0.4 hours | Intake flow controlled by demand and raw-water level |
| Coagulation and flocculation | 0.8 hours | Coagulant dosing varies by approximately ±12% |
| Sedimentation | 1.7 hours | Settled-water quality depends on raw-water conditions |
| Filtration | 0.9 hours per flow increment | Filter run length averages 38 hours |
| Backwash support activity | 0.75 hours | Each backwash takes 45 minutes |
| Disinfection | 0.5 hours | Residual chlorine ranges from 0.42 to 0.78 mg/L |
| Release, monitoring and transfer | 1.45 hours | Lab results and operational approval can delay release |
| Total process time | 6.5 hours | |
| Reservoir dwell | 9 hours | Storage used to balance demand and operations |
| Total lead time | 22 hours | Includes waiting, storage and information delays |
The 6.5-hour process time does not mean every litre waits for a single filter to complete its 38-hour run. Filters operate in parallel, allowing continuous flow while individual units approach headloss limits, enter backwash or return to service.
The map should also record quality and loss data:
- Filtrated-water turbidity target: under 0.3 NTU
- Current turbidity compliance: 96.8%
- Non-revenue water: 3%
- Unplanned downtime: 11.5 hours per month
- Primary downtime cause: valve faults
- Coagulant dosing variation: ±12%
- Chlorine residual target: 0.60 mg/L, with observed variation from 0.42 to 0.78 mg/L
4. Identify the eight wastes in a continuous water environment
The eight DOWNTIME wastes remain relevant even when the product flows continuously.
- Defects: Turbidity excursions, unstable chlorine residual or failed samples create rework, filter-to-waste activity and compliance risk.
- Overproduction: Treating beyond demand increases storage requirements, pumping energy and chemical consumption.
- Waiting: Operators may wait for laboratory results, maintenance approvals, replacement valves or regulatory sign-off.
- Non-utilised talent: Experienced operators spend time reconciling disconnected spreadsheets instead of improving control logic.
- Transportation: Unnecessary movement of chemical containers, samples or maintenance parts adds risk and handling time.
- Inventory: Excess chemical stock, spare parts or consumables tie up capital and can deteriorate or expire.
- Motion: Repeated manual trips to inspect gauges, collect readings or reset valves indicate opportunities for better visibility.
- Extra-processing: Over-sampling, duplicate data entry, repeated approvals and manual transcription add effort without increasing customer value.
Over-processing of sampling deserves particular attention. Sampling is essential, but sampling frequency, laboratory routing and reporting effort should be aligned with risk and process behaviour. A stable process may need a different sampling strategy from a plant experiencing frequent turbidity shifts.
5. Design the future-state map around control and flow
The future-state map should not simply shorten every time box. It should improve the system’s ability to detect, respond and learn.
Future-state design features
-
Level sensing and demand visibility
Add reliable reservoir and process-tank level sensing, connected to SCADA dashboards. This allows production to respond to actual demand rather than relying on delayed manual readings. -
Condition-based maintenance
Use valve position feedback, actuator-cycle counts, pressure trends and fault history to identify failure patterns before a valve interrupts flow. -
Standard work for backwash
Define trigger points using headloss, turbidity and run length. Standardise pre-checks, air scour, wash duration, return-to-service verification and post-backwash turbidity confirmation. -
SPC charting
Chart filtered-water turbidity and residual chlorine using time-ordered data. Operators can distinguish common-cause variation from a special cause such as a dosing fault, sensor drift or raw-water change. -
Integrated information flow
Connect SCADA alarms, laboratory results, work orders and regulatory reporting into a visual management system. The objective is not more data; it is faster and clearer decisions. -
Kaizen board
Display the top constraints, owners, due dates, countermeasures and verified results. A visible board creates a direct link between the value stream map and daily management.
Current state versus future state
| Measure | Current state | Future-state target |
|---|---|---|
| Total lead time | 22 hours | 15 hours |
| Process time | 6.5 hours | 5.8 hours |
| Value-added proportion | 29.5% | 38.7% |
| Turbidity compliance under 0.3 NTU | 96.8% | 99.5% |
| Non-revenue water | 3.0% | 2.0% |
| Unplanned downtime | 11.5 hours/month | 4.0 hours/month |
| Chlorine residual variation | 0.42–0.78 mg/L | 0.55–0.65 mg/L |
| Coagulant dosing variation | ±12% | ±5% |
These targets should be validated against local regulations, engineering constraints and actual customer requirements. The purpose of the future-state map is to establish a measurable operating model, not to create unsupported promises.
6. Sequence the kaizen work
A practical implementation sequence is:
-
Confirm the baseline : Operations and Quality, weeks 1–2
Validate flow, turbidity, chlorine, reservoir dwell, backwash frequency, NRW and downtime data. -
Stabilise measurement : SCADA and Laboratory, weeks 2–4
Check sensor calibration, align sampling timestamps and define one source of truth for reporting. -
Address valve faults : Maintenance, weeks 3–8
Rank valves by risk, review failure modes and introduce condition-based inspection for the highest-impact assets. -
Standardise backwash : Filtration Operations, weeks 4–7
Test trigger points, confirm the 45-minute standard and measure post-backwash recovery time and water use. -
Reduce dosing variation : Process Engineering, weeks 5–10
Analyse raw-water conditions, coagulant response and chlorine residual data. Use controlled trials before changing setpoints. -
Install SPC and visual escalation : Quality and Control Room, weeks 7–11
Establish control limits, escalation rules and rapid response routines for turbidity and chlorine signals. -
Reduce NRW : Distribution Team, weeks 8–16
Prioritise high-loss zones, validate meter accuracy and coordinate pressure management with customer-demand patterns. -
Control and sustain : Plant Manager, from week 12 onward
Review the future-state measures weekly, audit standard work and refresh the value stream map when process conditions change.
Value stream mapping makes the whole system visible: the physical journey of water, the movement of information and the decisions that determine whether customer value is delivered smoothly.
Build the capability to lead complex improvement projects: enrol in Lean Six Sigma Black Belt online training from Lean 6 Sigma Hub, accredited by CSSC and designed around practical, data-driven application.
Kaizen. Kai-Care. Kai-Done. ( Lean Six Sigma)
