Water Treatment: Separation, Physical Barriers and Disinfection

Running case. An intake experiences sudden storm turbidity, pesticide detections and low alkalinity during the wet season.

Unit product. Compare two treatment trains, perform a jar-test decision exercise and write an incident response sequence.

Unit orientation

This unit asks what can and cannot be concluded from observations in the running case. Begin by listing state variables and the time basis, identifying which quantities have been measured and which are assumptions. The engineering task is to distinguish evidence of transport, transformation, dilution and removal, and to prevent an untested model from becoming a safety claim. Compare at least two approaches with the same functional requirements, and include one sensitivity or uncertainty test in the written submission.

3.1 Coagulation and flocculation process selection

Colloids remain suspended because of particle size and surface electrochemistry. Coagulation changes stability, while flocculation promotes collisions and growth of settleable aggregates. Rapid mixing, chemical dose, pH, alkalinity and temperature interact. More coagulant does not always improve removal; overdosing can destabilize flocs, raise dissolved residuals or increase sludge costs.

Jar tests identify a useful operating window rather than a universal best dose. Compare clarified turbidity, filterability, pH change and residual metal with an untreated control. The engineer should avoid transferring a laboratory optimum directly to a full-scale plant without accounting for mixing energy, detention distribution and changing raw-water quality.

Governing relationship and units

Approximate mean velocity gradient G = √(P/(μV)).

Worked context

An operator increases alum dose to control a monsoon turbidity spike. The treated pH falls unexpectedly. Before further dosing, check raw-water alkalinity, actual chemical strength, pump calibration and sludge blanket performance. The immediate goal is stable barrier operation.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

3.2 Sedimentation and filtration hydraulics

In ideal discrete-particle sedimentation, a particle settles if its terminal velocity exceeds basin surface overflow rate. Real clarifiers deviate from this simple comparison because of density currents, floc growth, wind, sludge disturbance and short-circuiting. Filtration provides a different barrier: particles are captured within media or at surfaces, and head loss rises as solids accumulate.

Design reasoning starts with particle-size distribution and hydraulic loading, then checks residence time, solids storage, sludge withdrawal and peak flow. A filter can produce visually clear water while a bypass, seal failure or poorly managed backwash compromises the microbial barrier. Backwash timing and ripening after a run must be considered alongside pressure losses.

Governing relationship and units

Surface loading v₀=Q/A; Darcy head loss h=μLv/(ρgk).

Worked context

A clarifier carrying 4,000 m³/day over 200 m² has overflow velocity 20 m/day. Ideal particles settling at 30 m/day may be captured under assumed flow; particles at 5 m/day are not automatically captured by the same unit.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

3.3 Disinfection kinetics and multi-barrier control

Disinfection effectiveness depends on organism, disinfectant species, concentration, contact time, temperature, pH, suspended solids and hydraulic mixing. A single nominal CT product must not be equated with verified log removal. Residence-time distributions matter: the shortest effective contact time can be much less than tank volume divided by average flow.

Interpret residual decay as a process signal, not proof of pathogen inactivation. Use current jurisdictional guidance and validated contact-time methods for public-health decisions. A strong upstream particulate barrier can improve disinfection performance by reducing shielding; conversely, increasing chlorine to overcome poor filtration may create undesirable chemical byproducts.

Governing relationship and units

If residual follows C(t)=C₀e^(−kt), integrated concentration-time exposure = C₀(1−e^(−kt))/k.

Worked context

At initial 1.5 mg/L and 0.04 min⁻¹ decay, the predicted residual after 20 minutes is roughly 0.67 mg/L. This says nothing definitive about regulatory microbial credits without the organism, pH, temperature and hydraulics.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

3.4 Process trains, failures and water safety plans

No unit process should be judged solely by its average removal percentage. A water treatment train depends on combinations of source protection, particle removal, disinfection, pressure management and monitoring. If the source contains dissolved salts, ordinary clarification and sand filtration cannot reliably remove them; membrane processes or other separation approaches may be required, with reject-stream management.

Construct a hazard table with triggering event, early warning, preventive control, monitoring, corrective action and responsible person. Test plausible incidents: chemical stock-out, coagulant dosing fault, rapid turbidity increase, power interruption and cross-connection. A water safety plan is a living operating method rather than a once-off compliance document.

Governing relationship and units

C_out=C_in∏(1−η_i) is only justified if removal fractions are applicable, independent and defined for the same constituent.

Worked context

A pre-filter barrier removes 80% of particulate mass and a downstream filter removes 90% of the remaining mass, producing 98% total mass removal in an ideal two-stage calculation. That does not mean 98% pathogen protection.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

Synthesis and field decision

In the case scenario, compare two treatment trains, perform a jar-test decision exercise and write an incident response sequence. For each quantity used, identify its spatial representativeness, time basis, uncertainty and the responsible person who can confirm the value. Explicitly document how the plan changes during equipment failure or extreme conditions.

Unit assessment

1. For water treatment: separation, barriers and disinfection, define a control volume and list all stocks, transports and reaction terms with units.
2. Explain how different sampling times or locations could reverse the conclusion in this scenario: An intake experiences sudden storm turbidity, pesticide detections and low alkalinity during the wet season.
3. Choose two uncertainties and quantify which has the larger effect on the recommendation. Show a reproducible parameter table.
4. Use one of the four governing relationships to calculate a low, central and high case, then explain why a numerical result is not a permit limit.
5. Design an independent field or bench experiment that discriminates between a transport limitation and a reaction limitation.
6. Write a corrective-action plan for equipment failure or a surprising monitoring result; identify response time and independent verification.
7. Construct a single-page public explanation of the issue without concealing unresolved uncertainty.
8. Critique a proposal that reports percent removal but omits residual streams, peak-flow capacity and monitoring quality.

Assessment policy: A numerical answer without units, documented assumptions, and interpretation is incomplete. Explain what additional observation or regulation would be needed to authorize real-world action.

Unit 3: Water treatment: separation, barriers and disinfection

Running case. An intake experiences sudden storm turbidity, pesticide detections and low alkalinity during the wet season.

Unit product. Compare two treatment trains, perform a jar-test decision exercise and write an incident response sequence.

Unit orientation

This unit asks what can and cannot be concluded from observations in the running case. Begin by listing state variables and the time basis, identifying which quantities have been measured and which are assumptions. The engineering task is to distinguish evidence of transport, transformation, dilution and removal, and to prevent an untested model from becoming a safety claim. Compare at least two approaches with the same functional requirements, and include one sensitivity or uncertainty test in the written submission.

3.1 Coagulation and flocculation process selection

Colloids remain suspended because of particle size and surface electrochemistry. Coagulation changes stability, while flocculation promotes collisions and growth of settleable aggregates. Rapid mixing, chemical dose, pH, alkalinity and temperature interact. More coagulant does not always improve removal; overdosing can destabilize flocs, raise dissolved residuals or increase sludge costs.

Jar tests identify a useful operating window rather than a universal best dose. Compare clarified turbidity, filterability, pH change and residual metal with an untreated control. The engineer should avoid transferring a laboratory optimum directly to a full-scale plant without accounting for mixing energy, detention distribution and changing raw-water quality.

Governing relationship and units

Approximate mean velocity gradient G = √(P/(μV)).

Worked context

An operator increases alum dose to control a monsoon turbidity spike. The treated pH falls unexpectedly. Before further dosing, check raw-water alkalinity, actual chemical strength, pump calibration and sludge blanket performance. The immediate goal is stable barrier operation.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

3.2 Sedimentation and filtration hydraulics

In ideal discrete-particle sedimentation, a particle settles if its terminal velocity exceeds basin surface overflow rate. Real clarifiers deviate from this simple comparison because of density currents, floc growth, wind, sludge disturbance and short-circuiting. Filtration provides a different barrier: particles are captured within media or at surfaces, and head loss rises as solids accumulate.

Design reasoning starts with particle-size distribution and hydraulic loading, then checks residence time, solids storage, sludge withdrawal and peak flow. A filter can produce visually clear water while a bypass, seal failure or poorly managed backwash compromises the microbial barrier. Backwash timing and ripening after a run must be considered alongside pressure losses.

Governing relationship and units

Surface loading v₀=Q/A; Darcy head loss h=μLv/(ρgk).

Worked context

A clarifier carrying 4,000 m³/day over 200 m² has overflow velocity 20 m/day. Ideal particles settling at 30 m/day may be captured under assumed flow; particles at 5 m/day are not automatically captured by the same unit.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

3.3 Disinfection kinetics and multi-barrier control

Disinfection effectiveness depends on organism, disinfectant species, concentration, contact time, temperature, pH, suspended solids and hydraulic mixing. A single nominal CT product must not be equated with verified log removal. Residence-time distributions matter: the shortest effective contact time can be much less than tank volume divided by average flow.

Interpret residual decay as a process signal, not proof of pathogen inactivation. Use current jurisdictional guidance and validated contact-time methods for public-health decisions. A strong upstream particulate barrier can improve disinfection performance by reducing shielding; conversely, increasing chlorine to overcome poor filtration may create undesirable chemical byproducts.

Governing relationship and units

If residual follows C(t)=C₀e^(−kt), integrated concentration-time exposure = C₀(1−e^(−kt))/k.

Worked context

At initial 1.5 mg/L and 0.04 min⁻¹ decay, the predicted residual after 20 minutes is roughly 0.67 mg/L. This says nothing definitive about regulatory microbial credits without the organism, pH, temperature and hydraulics.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

3.4 Process trains, failures and water safety plans

No unit process should be judged solely by its average removal percentage. A water treatment train depends on combinations of source protection, particle removal, disinfection, pressure management and monitoring. If the source contains dissolved salts, ordinary clarification and sand filtration cannot reliably remove them; membrane processes or other separation approaches may be required, with reject-stream management.

Construct a hazard table with triggering event, early warning, preventive control, monitoring, corrective action and responsible person. Test plausible incidents: chemical stock-out, coagulant dosing fault, rapid turbidity increase, power interruption and cross-connection. A water safety plan is a living operating method rather than a once-off compliance document.

Governing relationship and units

C_out=C_in∏(1−η_i) is only justified if removal fractions are applicable, independent and defined for the same constituent.

Worked context

A pre-filter barrier removes 80% of particulate mass and a downstream filter removes 90% of the remaining mass, producing 98% total mass removal in an ideal two-stage calculation. That does not mean 98% pathogen protection.

Reasoning exercise. Using the running case, make a two-column table: available observations and missing evidence. State a plausible parameter range with units, identify one observation capable of disproving the principal assumption, and explain how the engineering action would change if a measured input were doubled. Do not treat the worked illustrative numbers as calibration data.

Model critique. Check dimensional consistency before calculation. If data come from different dates, locations, temperature bases or chemical definitions, reconciliation is required before drawing a conclusion. Report the sensitivity of the decision to uncertain inputs and distinguish the physical result from any legal or public-health judgment.

Synthesis and field decision

In the case scenario, compare two treatment trains, perform a jar-test decision exercise and write an incident response sequence. For each quantity used, identify its spatial representativeness, time basis, uncertainty and the responsible person who can confirm the value. Explicitly document how the plan changes during equipment failure or extreme conditions.

Unit assessment

1. For water treatment: separation, barriers and disinfection, define a control volume and list all stocks, transports and reaction terms with units.
2. Explain how different sampling times or locations could reverse the conclusion in this scenario: An intake experiences sudden storm turbidity, pesticide detections and low alkalinity during the wet season.
3. Choose two uncertainties and quantify which has the larger effect on the recommendation. Show a reproducible parameter table.
4. Use one of the four governing relationships to calculate a low, central and high case, then explain why a numerical result is not a permit limit.
5. Design an independent field or bench experiment that discriminates between a transport limitation and a reaction limitation.
6. Write a corrective-action plan for equipment failure or a surprising monitoring result; identify response time and independent verification.
7. Construct a single-page public explanation of the issue without concealing unresolved uncertainty.
8. Critique a proposal that reports percent removal but omits residual streams, peak-flow capacity and monitoring quality.

Assessment policy: A numerical answer without units, documented assumptions, and interpretation is incomplete. Explain what additional observation or regulation would be needed to authorize real-world action.