Operational Data, Trends & Performance Review

Water and wastewater operators collect large amounts of information every day. Flow, pressure, chemical feed, turbidity, chlorine residual, dissolved oxygen, sludge levels, laboratory results, equipment run times, and many other measurements describe how the system is performing.

Individual readings are important, but the greatest value often comes from reviewing those readings over time. Trend analysis helps operators recognize gradual changes, identify developing problems, evaluate the effect of process adjustments, and determine whether performance is improving or deteriorating.

Data Should Support Decisions

Collecting data has little value if nobody reviews it.

Operational data should help answer practical questions such as:

• Is treatment stable?
• Is water quality improving or declining?
• Is chemical demand changing?
• Is equipment performance deteriorating?
• Are flows increasing?
• Are process adjustments producing the expected result?
• Is the system moving toward a permit or regulatory limit?

The goal is to convert measurements into useful operating information.

A Single Reading Versus a Trend

A single measurement tells the operator what was observed at one moment.

A trend shows how the measurement is changing over time.

For example, a chlorine residual of 1.2 mg/L may be acceptable at one system. However, if recent values were:

• 1.8 mg/L;
• 1.6 mg/L;
• 1.5 mg/L;
• 1.3 mg/L;
• 1.2 mg/L;

the downward trend may require investigation even before the residual reaches an unacceptable level.

Establish Normal Operating Ranges

Operators need a reference point for deciding whether a value is normal.

A normal operating range may come from:

• permit requirements;
• regulatory standards;
• manufacturer recommendations;
• facility SOPs;
• historical plant performance;
• engineering design;
• experienced operator knowledge.

Normal operating ranges should not be confused with regulatory limits. A facility may use an internal target that provides a safety margin before a regulatory limit is reached.

Operating Target Versus Regulatory Limit

An operating target is a preferred value or range used to maintain stable treatment.

A regulatory limit is a legal requirement established by a permit or regulation.

For example, a facility may maintain an internal process target that is more conservative than the maximum allowed value.

Exceeding the internal target may signal the need for corrective action without yet creating a regulatory violation.

Use Historical Data

Historical records provide context for current conditions.

Useful comparisons include:

• today versus yesterday;
• this week versus last week;
• this month versus the same month last year;
• dry weather versus wet weather;
• summer versus winter;
• before and after a process change;
• before and after equipment maintenance.

Historical comparison helps distinguish normal variation from unusual change.

Daily Review

Operators should review important process data frequently enough to recognize developing problems.

A daily review may include:

• flow;
• tank or reservoir levels;
• pressure;
• chemical-feed rates;
• chemical inventory;
• disinfectant residual;
• turbidity;
• pH;
• dissolved oxygen;
• laboratory results;
• equipment status;
• alarms.

The specific data depend on the facility.

Wastewater Trend Examples

Wastewater operators commonly review trends involving:

• influent flow;
• influent BOD or COD;
• influent TSS;
• aeration-basin dissolved oxygen;
• MLSS and MLVSS;
• SVI;
• sludge blanket depth;
• RAS rate;
• wasting rate;
• ammonia;
• effluent BOD;
• effluent TSS;
• disinfection performance.

Several related parameters should often be reviewed together rather than independently.

Drinking Water Trend Examples

Drinking water operators commonly review trends involving:

• source-water turbidity;
• raw-water quality;
• coagulant dose;
• pH;
• filter run time;
• filter effluent turbidity;
• chlorine dose;
• chlorine residual;
• distribution-system residual;
• tank levels;
• system pressure;
• customer complaints.

Changes in one parameter may help explain changes in another.

Look for Relationships

Trend analysis is especially useful when operators compare related variables.

Examples include:

• flow versus chemical use;
• source turbidity versus coagulant dose;
• chemical dose versus finished-water residual;
• MLSS versus wasting rate;
• dissolved oxygen versus blower output;
• rainfall versus wastewater influent flow;
• pump run time versus system demand.

Relationships help operators understand cause and effect.

Example: Chemical Dose and Residual

Suppose chlorine dose remains constant while finished-water residual gradually decreases.

Possible explanations include:

• increased chlorine demand;
• change in source-water quality;
• higher flow;
• chemical-feed problem;
• instrument or test error.

Reviewing dose, flow, residual, source conditions, and equipment status together gives more information than reviewing residual alone.

Example: Wastewater Flow and Treatment Performance

Suppose influent flow increases sharply during rainfall and effluent suspended solids also begin to increase.

The operator might review:

• rainfall;
• influent flow;
• clarifier loading;
• sludge blanket depth;
• RAS rate;
• effluent TSS;
• collection-system conditions.

The relationship may indicate hydraulic stress caused by wet-weather flow.

Recognize Gradual Change

Many failures develop gradually.

Examples include:

• pump capacity slowly declining;
• filter runs becoming shorter;
• chemical demand increasing;
• sludge blanket rising;
• dissolved oxygen becoming harder to maintain;
• distribution pressure slowly decreasing.

Operators who review trends may recognize these problems before an alarm or violation occurs.

Rate of Change

The speed of change can be as important as the actual value.

A small change over several months may indicate slow deterioration.

The same change over 30 minutes may indicate an active process upset or equipment failure.

Operators should consider:

• magnitude of change;
• direction of change;
• speed of change;
• duration of change.

Use Graphs When Helpful

Graphs often make trends easier to see than tables of numbers.

A useful graph may show:

• parameter on the vertical axis;
• time on the horizontal axis;
• target range;
• important operating events;
• process changes.

Graphing can make gradual drift, cycles, and sudden changes more obvious.

Choose the Correct Time Scale

The same data can look different depending on the time scale.

Operators may review:

• minutes or hours for an active process upset;
• days for short-term process control;
• weeks for maintenance or biological-process trends;
• months or years for seasonal and long-term performance.

Use a time scale appropriate for the question being investigated.

Average Values

Averages can simplify large amounts of data.

For example:

Average = Sum of Values ÷ Number of Values

If daily flows for five days are:

• 1.20 MGD;
• 1.35 MGD;
• 1.25 MGD;
• 1.50 MGD;
• 1.45 MGD;

the average is:

(1.20 + 1.35 + 1.25 + 1.50 + 1.45) ÷ 5 = 1.35 MGD

An average can be useful, but it can also hide short-term high or low values.

Do Not Rely Only on Averages

Two systems can have the same average but very different operating behavior.

For example, one process may remain close to the average every day while another alternates between very high and very low values.

Operators should review:

• average;
• minimum;
• maximum;
• range;
• individual abnormal values.

Percent Change

Percent change can help describe how much a value has increased or decreased.

A basic formula is:

Percent Change = (New Value - Old Value) ÷ Old Value × 100

If average chemical use increases from 400 lb/day to 500 lb/day:

(500 - 400) ÷ 400 × 100 = 25%

Chemical usage increased by 25%.

The next question is why.

Key Performance Indicators

A Key Performance Indicator, or KPI, is a selected measurement used to summarize important system performance.

Useful operator-focused KPIs may include:

• permit compliance rate;
• finished-water turbidity;
• chemical use per unit of flow;
• energy use per unit of treatment;
• filter run time;
• pump efficiency;
• equipment downtime;
• number of sanitary sewer overflows;
• customer complaints;
• preventive-maintenance completion.

A KPI should measure something operationally useful rather than simply something easy to count.

Normalize Data When Appropriate

Raw chemical use may increase simply because flow increased.

To make a fair comparison, operators may normalize usage.

For example:

Chemical Use per MG = Chemical Used ÷ Flow Treated

If a plant used 600 lb of chemical while treating 3 MG:

600 lb ÷ 3 MG = 200 lb/MG

This value can be compared with other days even when total flow differs.

Compare Actual Performance with Targets

A useful review compares current values with:

• target;
• normal range;
• previous period;
• regulatory limit;
• design value where relevant.

This makes it easier to determine whether performance is acceptable or moving in the wrong direction.

Alarm Data

Alarm history can also provide useful performance information.

Repeated alarms may indicate:

• unstable process conditions;
• poor alarm setpoints;
• instrument problems;
• equipment deterioration;
• staffing or response issues.

An alarm that occurs frequently should not simply become background noise.

Avoid Alarm Fatigue

When operators receive too many unnecessary alarms, important alarms may be overlooked.

Facilities should review repeated or nuisance alarms and determine whether:

• setpoints are appropriate;
• equipment requires repair;
• instrument calibration is needed;
• process conditions should be corrected.

Alarm review is part of performance review.

Data Quality Comes First

Trend analysis is only as reliable as the underlying data.

Before reacting to an unusual value, consider whether the reading could be affected by:

• instrument calibration;
• sample error;
• laboratory error;
• incorrect units;
• data-entry error;
• sensor fouling;
• communication failure.

An unexpected result should be investigated, not automatically discarded.

Verify Unexpected Values

If a reading is inconsistent with all other information, verify it when appropriate.

Verification may include:

• repeat measurement;
• instrument check;
• calibration check;
• comparison with another instrument;
• review of sample handling;
• field inspection.

Do not change or delete a valid result simply because it is inconvenient.

Distinguish Signal from Noise

Normal systems show some variation.

Operators should distinguish ordinary variation from meaningful process change.

Useful questions include:

• Is the change larger than normal variation?
• Is it continuing in one direction?
• Did another parameter change at the same time?
• Did equipment or weather conditions change?
• Was a process adjustment recently made?

Evaluate Process Adjustments

When an operator changes a process, follow-up data should determine whether the change worked.

A practical sequence is:

1. Identify the problem.
2. Record the initial condition.
3. Make one appropriate adjustment.
4. Allow adequate process response time.
5. Measure the result.
6. Compare with the expected response.
7. Make another adjustment only if needed.

Changing several variables at once can make it difficult to determine which change produced the result.

Consider Process Response Time

Not every process responds immediately.

Examples include:

• distribution-system changes;
• large storage tanks;
• biological wastewater processes;
• chemical mixing systems;
• solids inventory changes.

Operators should understand the expected response time before deciding whether an adjustment failed.

Review Performance After Maintenance

Maintenance should improve or restore equipment performance.

After maintenance, compare:

• flow;
• pressure;
• power use;
• vibration;
• run time;
• alarm frequency;
• other relevant measurements.

This helps confirm whether the repair produced the expected improvement.

Use Data for Preventive Maintenance

Operating trends can identify equipment deterioration before failure.

Examples include:

• increasing motor current;
• decreasing pump flow;
• increasing vibration;
• increasing bearing temperature;
• more frequent chemical-pump calibration drift.

These conditions may indicate the need for inspection or maintenance.

Customer Complaints as Data

Customer complaints can be treated as another operational data source.

Trends may involve:

• taste and odor;
• low pressure;
• discoloration;
• sewer backups;
• odor complaints;
• service interruptions.

Patterns by location or time can reveal problems that are not obvious from plant data alone.

Review Performance Regularly

Performance review should occur at appropriate intervals.

Examples include:

• shift review;
• daily review;
• weekly operations meeting;
• monthly compliance review;
• annual performance assessment.

Different review periods serve different purposes.

Daily Performance Review

Daily review focuses on immediate operating conditions.

Questions may include:

• Is the process stable?
• Are critical values within normal operating ranges?
• Did any alarms occur?
• Did equipment fail?
• Are follow-up actions required?

Weekly or Monthly Review

Longer-term review can identify trends that are difficult to see during one shift.

Questions may include:

• Is chemical use increasing?
• Is energy use increasing?
• Are filter runs becoming shorter?
• Are permit values moving closer to limits?
• Is maintenance backlog increasing?
• Are customer complaints becoming more frequent?

Document Conclusions and Actions

Performance review should lead to action when appropriate.

A useful review record may identify:

• trend observed;
• possible cause;
• risk;
• recommended action;
• responsible person;
• follow-up date.

Recognizing a trend without assigning follow-up may allow the problem to continue.

Do Not Manipulate Data to Improve Appearance

Operational review must use actual data.

Operators should never:

• remove unfavorable values from a trend without a valid documented reason;
• change units to make performance appear better;
• average data in a way that hides a regulatory exceedance;
• replace actual data with estimated values without clear identification;
• ignore abnormal results because they do not match expectations.

Accurate data are necessary for both good operation and regulatory integrity.

Common Data Review Mistakes

• Collecting data without reviewing it.
• Looking only at one reading instead of the trend.
• Confusing internal operating targets with regulatory limits.
• Ignoring gradual changes because values remain within limits.
• Reviewing one parameter without related process data.
• Using averages that hide important maximum or minimum values.
• Reacting to an unusual result without checking data quality.
• Discarding unexpected data without investigation.
• Changing several process variables at once.
• Expecting an immediate response from a slow process.
• Ignoring repeated alarms.
• Failing to document follow-up actions.

A Practical Performance Review Sequence

1. Confirm that the data are complete and reliable.
2. Compare current values with normal operating ranges.
3. Review recent trends.
4. Compare related parameters.
5. Identify abnormal changes or movement toward limits.
6. Consider equipment, weather, flow, and process changes.
7. Verify unexpected values when appropriate.
8. Determine whether corrective action is needed.
9. Make one controlled adjustment when practical.
10. Allow appropriate process response time.
11. Measure the result.
12. Document conclusions and follow-up.

What to Remember for the Exam

• Operational data are most useful when they support process-control decisions.
• A single value shows one moment; a trend shows direction over time.
• Operators should know normal operating ranges for important parameters.
• Internal operating targets may be more conservative than regulatory limits.
• Historical data help distinguish normal variation from abnormal change.
• Related parameters should often be reviewed together.
• Gradual trends can reveal developing problems before alarms or violations occur.
• The rate of change can be as important as the actual value.
• Graphs can make trends easier to recognize.
• Averages are useful but can hide maximums, minimums, or short-term problems.
• Percent change can help describe how much performance has changed.
• Normalizing values, such as chemical use per MG treated, can improve comparisons.
• Data quality should be checked before making major decisions from unusual values.
• Unexpected values should be investigated, not automatically discarded.
• After a process adjustment, operators should evaluate whether the expected response occurred.
• Slow processes require enough time before the effect of an adjustment can be judged.
• Repeated alarms and customer complaints can provide useful performance information.
• Performance review should lead to documented corrective action when needed.
• Never alter, omit, or manipulate valid data simply to make performance appear better.