16.5 Preventive Maintenance, CMMS & Asset Management
Key Takeaways
- Equipment Installation and Repair and Troubleshoot and Repair Pumps and Motors are named sub-topics in the SWRCB distribution Equipment category, and Maintenance Plan is named in the regulations category.
- Maintenance strategies range from reactive to preventive to predictive to reliability-centered, and the right mix depends on the consequence of failure.
- 22 CCR 64600 requires a public water system to have an operations and maintenance plan.
- Predictive tools include vibration analysis, infrared thermography, oil analysis, motor circuit analysis, and ultrasonic detection.
- Asset management ranks work by criticality, the product of likelihood of failure and consequence of failure, rather than by age alone.
Maintenance Is a Regulatory Requirement
22 CCR 64600 requires a public water system to establish a Water System Operations and Maintenance Plan, and the distribution blueprint lists Maintenance Plan as a named sub-topic under Drinking Water Regulations / Management / Safety - the category worth 35 questions at D4 and 45 at D5. On the wastewater side, 23 CCR 3701(c)(4) puts supervision and management responsibilities including operator training and budget control into the Grade IV blueprint, and a maintenance program sits squarely inside that.
Maintenance is also a sanitary survey element: "system management and operation" is one of the eight elements, and an inability to demonstrate a maintenance program is a documented deficiency.
The Four Strategies
| Strategy | Description | Appropriate when |
|---|---|---|
| Reactive (run to failure) | Fix it when it breaks | The asset is cheap, redundant, and non-critical, and failure has no downstream consequence - a light fixture, a spare hose |
| Preventive (time or run-hour based) | Service on a fixed interval | Failure modes are age-related and the interval is known - oil changes, filter changes, valve exercising, generator exercise |
| Predictive (condition-based) | Measure condition and act on the trend | The asset is expensive or critical and has a measurable degradation signal - large motors, blowers, high service pumps |
| Reliability-centered (RCM) | Analyze each failure mode and choose the strategy that addresses it | High-consequence systems where a structured analysis pays for itself |
[!IMPORTANT] Not everything should be on a PM schedule. Over-maintaining is a real failure mode: every intervention introduces a chance of infant mortality from a mis-assembly, a contaminated lubricant, or a disturbed connection. The question is always what failure mode does this task prevent? If the answer is "none, we've always done it," the task is a candidate for elimination.
Designing a Preventive Maintenance Program
- Build an asset register. Every asset with a unique ID, location, make, model, serial number, size or rating, installation date, and parent system.
- Rank by criticality.
Consequence includes public health, permit compliance, customers affected, safety, environmental impact, and repair cost. A small chemical feed pump with no standby at a plant with no other disinfection can be more critical than a large pump with three redundant units.
- Define tasks and intervals from the manufacturer's requirements, industry practice, and your own failure history. Intervals may be calendar-based, run-hour based, or condition-based.
- Write the task so anyone can do it - specific steps, torque values, lubricant type and quantity, acceptance criteria, and required PPE and lockout points.
- Schedule and level the workload so PM does not all land in one week.
- Close the loop. Record what was found, not just that the task was done. "Found coupling insert worn, replaced" is data; "PM complete" is not.
Typical Intervals
| Asset | Typical PM |
|---|---|
| Pumps | Weekly visual and vibration check; quarterly lubrication; annual alignment verification; seal and packing per condition |
| Motors | Annual megger insulation test; thermography under load; bearing lubrication per schedule |
| Blowers | Inlet filter differential monitoring; lubrication; annual relief valve test |
| Distribution valves | Exercise annually for critical/boundary, every 1-3 years for the rest |
| Hydrants | Annual inspection and operational test |
| Generators | Weekly no-load exercise; annual load bank test |
| Backflow assemblies | Annual certified test |
| Chemical feed pumps | Monthly drawdown calibration; diaphragm/tube replacement per hours |
| Analyzers | Daily to weekly grab-sample verification; scheduled reagent and electrode service |
| Storage tanks | Routine exterior/vent/hatch inspection; interior inspection every 3-5 years |
Predictive Technologies
| Technology | Detects |
|---|---|
| Vibration analysis | Imbalance (1x running speed), misalignment (2x, axial), bearing defects (high-frequency bearing tones), looseness, cavitation (broadband random), impeller damage. The single most valuable rotating-equipment tool |
| Infrared thermography | Hot electrical connections, overloaded conductors, failing breakers, bearing overheating, steam and insulation losses. A loose lug shows up as a hot spot long before it fails |
| Oil analysis | Wear metals, water and glycol contamination, viscosity change, additive depletion |
| Motor circuit analysis / megger | Insulation deterioration, winding faults, rotor bar problems |
| Ultrasonic detection | Compressed air leaks (pure wasted energy), steam trap failures, early bearing distress, electrical arcing |
| Performance trending | Pump head-capacity versus the original curve; blower pressure at a reference airflow; specific capacity of a well; C-factor of a main |
[!TIP] Performance trending is free predictive maintenance. A pump whose delivered flow at a given head has fallen 15 percent has worn wear rings or a damaged impeller, and you learned it from SCADA data you already had. A blower whose discharge pressure at constant airflow has risen has fouled diffusers. A well whose specific capacity has fallen has a fouled screen. None of that requires a new instrument.
CMMS
A computerized maintenance management system turns maintenance from memory into a managed process. Core functions: the asset register, PM scheduling and automatic work order generation, corrective work orders, work history by asset, parts inventory linked to assets, labor and cost tracking, and reporting.
Metrics worth watching:
- PM compliance - percentage of scheduled PMs completed on time
- PM to corrective ratio - a healthy program trends toward more planned than reactive work
- Mean time between failures (MTBF) by asset class
- Backlog in crew-weeks - a backlog that grows steadily means the program is under-resourced
- Emergency work percentage - the clearest signal of program health
Work history is the real payoff. When the same pump has been rebuilt four times in three years, the CMMS is the record that justifies replacing it rather than rebuilding it a fifth time - and that record is what turns a maintenance opinion into a capital budget request.
Spare Parts
Stock decisions follow criticality and lead time:
| Category | Policy |
|---|---|
| Critical spares - single points of failure with long lead times | Stock, even at high carrying cost. A custom pump shaft with a 20-week lead time on a non-redundant service is a stocking decision, not an inventory optimization problem |
| Consumables - gaskets, seals, filters, lubricants, belts | Min/max reorder points |
| Common parts - standard bearings, fasteners | Vendor-managed or local supply |
| Standardization | Specifying fewer pump, motor, and valve models across the utility cuts inventory dramatically and increases the odds that a spare fits |
Asset Management
Asset management links maintenance to capital planning by answering five questions:
- What do I own? The asset register and the distribution system map.
- What is it worth? Replacement cost, and the value of the service it provides.
- What is its condition and remaining life? Condition assessment, break history, performance trending.
- What is its criticality? Consequence of failure.
- What is my best investment strategy? The lowest life-cycle cost that maintains the required level of service at an acceptable risk.
The output is a capital improvement plan driven by risk rather than by age, and a rate structure that funds renewal and replacement rather than deferring it. In California, a utility that cannot demonstrate this is a utility with a technical, managerial, and financial capacity problem, and TMF capacity is a permit issue - which is how maintenance planning connects back to the Division of Drinking Water.
A small non-redundant chemical feed pump provides the only disinfection at a groundwater plant, while a large high-service pump is one of four identical units. Which is more critical from an asset management standpoint?
Vibration analysis of a pump shows a strong peak at twice the running speed with significant axial vibration. What does this pattern most commonly indicate?
A utility performs preventive maintenance on every asset in its inventory on a fixed calendar interval regardless of criticality. What is the principal weakness of this approach?