Monroe L. Weber-Shirk S chool of Civil and Environmental Engineering Technology Selection Reflections Getting rid of all the muck Biggest bang for the.

Monroe L. Weber-Shirk

School of Civil and

Environmental Engineering

Technology Selection ReflectionsTechnology Selection Reflections

Getting rid of all the muck

Biggest bang for the buck

Reliability, no need for luck

Selecting a Treatment ProcessSelecting a Treatment Process

Water characteristics

Resources (Capacities)

Institutional

Economic

Labor force

Infrastructure

Decision

Input Algorithm Output

TreatmentProcess

Education

Scale

Decision Quality as f(Data Quantity)Decision Quality as f(Data Quantity)T

reat

men

t Cho

ice

Dec

isio

n Q

uali

ty

Amount of Data

optimal

More data, but no design change!

How could you increase the y intercept? ____________Better default!

How could you increase the slope? _________________Identify critical data!

Optimal Water Treatment DecisionOptimal Water Treatment Decision

SustainableImprovement in

Public health (risk reduction)Labor savingsIndividual and community empowerment

At a cost/benefit ratio that is commensurate with competing expenditures and interventions

An Optimization Problem with Many Options

An Optimization Problem with Many Options

TechnologyWater sourcesWater treatment processesWater storageWater distributionSeparate drinking water from other uses (bottled water)

Scale (household to municipal) Staging (order of implementation)

Sustainable Staged Space

Data QualityData Quality

Many of the choices are discrete (either process A or B or C)

Thus there are regions with additional data that don’t cause any improvement in design

How can we choose which data to gather to maximize the rate of approach to the optimal design?

We will return to this question after we review our options

What are our Choices?Clean Water Combos

What are our Choices?Clean Water Combos

Water SourceScale, type, characteristics

TreatmentScale, capacity, processes, automation

StorageScale, capacity

Distribution resolutionScale, capacity

1 10 100 1k 10kmeters

100k1 10 100 1k 10kmeters

100k

1 10 100 1k 10kpeople

100k1 10 100 1k 10kpeople

100k

1 10 100 1k 10kpeople

100k1 10 100 1k 10kpeople

100k

1 10 100 1k 10kmeters

100k1 10 100 1k 10kmeters

100k

Ithaca

1 10 100 1kLiters

person day

1 10 100 1kLiters

person daycapacity

Water Characteristics: SourceWater Characteristics: Source

RainTreat as if it were surface water

GroundwaterIf “under the influence,” then treat as if it were

surface water

SurfaceOcean

Water Treatment ObjectivesWater Treatment Objectives

Particle removalGet turbidity below

30 NTU (WHO limit for disinfection only treatments)

5 NTU (Particle removal technologies should exceed this goal)

Pathogen inactivation/removal

Hazardous chemical removalNaturally occurring

ArsenicFluorideNitrate/nitrite

Anthropogenic contamination

Microbiological Safety Chemical Safety

WHO is working on guidance for these contaminants

1

2

Particle Removal: Big ScaleParticle Removal: Big Scale

1000

NTU

1

10

100

SSF Contact Direct Conventional

1 10 100 1k 10kpeople

100k1 10 100 1k 10kpeople

100k

Approximate turbidity range

Operator Skill*low

mediumadvanced

*EPA’s opinion, not WHO’s opinion!

Particle Removal: Small ScaleParticle Removal: Small Scale

1000

NTU

1

10

100

SSF

1 10 100 1k 10kpeople

100k1 10 100 1k 10kpeople

100k

PuR Cartridge BagFloc/Sed

Consumables?

Pot Candle

sand? alum PuR filters

$0 $10$

cap yr$1

WHO on Particle Removal for POUWHO on Particle Removal for POU

There is a need to investigate, characterize and implement physical and physical-chemical technologies for practical and low cost pre-treatment

Some physical or physical-chemical methods may be highly effective for treatment of stored household water on their own. (i.e., won’t need disinfection)

Particle removal technologies include: Settling or plain sedimentation Fiber, cloth or membrane filters Granular media filters Slow sand filter

WHO on SSF as POUWHO on SSF as POU

Slow sand filtration is the least likely to be sustainable at the household level. the preferred filter designs and installations often are larger and

capable of treating more water than needed by individual households

because of their relatively large size (surface area) and the needs for

proper construction and operation, regular maintenance (especially sand scraping, replacement and

cleaning) by trained individuals. Such demands for achieving good performance are

unrealistic because they are beyond the capacities and capabilities of most households

Need a good small-scale design!

Need a simple cleaning technique!

What was WHO thinking about SSF?

What was WHO thinking about SSF?

How much water will this system produce?_____ m/hr_____ m/d_____ m3

Why won’t this system work well?

0.45 m

0.12.4

0.38

SSF Design Flaws…SSF Design Flaws…

Scour when head loss is lowRequires a hill side

3 200 L drumsSiphon risk-Top layer of sand can dewater if supply water stops or if head loss is low

Expensive

Takes up lots of space

Flow control (“floating weir”)

Can’t handle much head loss

Flow Control FailureFlow Control Failure

A floating weir (that can be made of a bowl, two small tubes and a hose) in the supply tank maintains a constant flow of water to the top of the filter tank

Environmental Health Project (WASH ) concludes that the close attention and frequent adjustment required to operate demonstration models has resulted in early abandonment

Why doesn’t this work well?Why doesn’t this work well?

Where is constant head? Where is head loss

element? How is flow adjusted? What is the role of the

nylon string? What happens when you

add a pebble? How flexible is a rubber

tube?

The Proctor and Gamble Solution: PuR

The Proctor and Gamble Solution: PuR

The PuR product uses ferric sulfate, bentonite, sodium carbonate, chitosan, polyacrylamide, potassium permanganate, and calcium hypochlorite

A small sachet of powdered product visibly separates the cleaned water from the murky masses

Initial efforts are underway to develop a sustainable market-based approach for delivery and to learn how to best make POU products available. Three separate complementary models are being explored: a social model led by non-profit organizations a commercial model led by the private sector an emergency relief model led by relief organization

One small sachet, costing about US $0.10 in the commercial model, will treat 10 liters of water (enough drinking water for an average family for two days)

PuR: DirectionsPuR: Directions

Add 1 sachet to 10 litres of water and stir to begin process of separating the cleaned water from the murky masses

Stir water for 5 minutes until clear Filter water through a cloth and dispose of

separated floc in the latrine Let clear water stand for 20 minutes to allow for

complete disinfection Store in a suitable container to prevent

recontamination

No sedimentation?

PuR: Microorganisms andArsenic Removal

PuR is expected to provide excellent disinfection (>7-log bacterial, >4-log viral and >3-log parasite reductions) across a variety of water types and under conditions that stress less effective purification products including solar or chlorine treatment alone

No E. coli were detected post-treatment in any of 320 samples of drinking water sources collected in developing countries

The POU treatment was also effective in removing arsenic from water artificially contaminated with arsenic and from water with naturally occurring arsenic contamination

In Bangladesh tests, arsenic decreased by a mean of (85%) 88% of treated samples were <50 ppb

PuR Turbidity RangePuR Turbidity Range

Turbidities in the samples were reduced significantly, pre-treatment ranged from 0 to 1850 NTU (mean 19 NTU) and final values were generally less than 1 NTU (average 0.25 NTU).

The highest final turbidity observed was 3.2 NTU for a water source whose starting turbidity had 1850 NTU

PuR CritiquePuR Critique

This is not sustainable or in the interests of people in rural areas.

It becomes a product that has to be purchased on a regular basis from a foreign country.

I think the analogy to the scandalous infant formula problems of a couple of decades ago should be kept in mind where people were encouraged to abandon breast feeding in favor of a foreign infant formula.

Getting people “hooked” on a product that will require as much as 10% of their income instead of trying to develop sustainable solutions that don’t have recurrent cost and that the villagers have control over is exploitive in the worst of ways

--Humphrey Blackburn* *Okay, he designs and sells slow sand filters…

Particle Removal: Small ScaleParticle Removal: Small Scale

1000

NTU

1

10

100

SSF

1 10 100 1k 10kpeople

100k1 10 100 1k 10kpeople

100k

PuR Cartridge BagFloc/Sed

Consumables?

Pot Candle

sand? alum PuR filters

$0 $10$

cap yr$1

Minimal Data Requirements for Surface Water Treatment

Minimal Data Requirements for Surface Water Treatment

What would you need to know before you would be willing to recommend a water treatment technology for a community of 250 that is currently relying on an untreated surface water source?

Minimal Data…Minimal Data…

Turbidity Pathogens Chemicals

Determine if naturally occurring contaminants are present in region

Assess watershed exposure risk to agricultural and industrial contamination

Economic, Institutional, Educational Capacity

Will determine treatment technology

Assume pathogens are present!

The Choice of ScaleThe Choice of Scale

My long held assumption that only centralized systems made sense

Remember creativity: vary parameters over the full range of possibilitiesVary number of customers per treatment plant!

Are there situations where decentralized is better?

1 100k

Centralized Models in the Global North

Centralized Models in the Global North

Centralized (Municipal) Water source (possibly multiple sources) Treatment (possibly multiple facilities) Storage (usually multiple tanks in sprawling communities) Distribution (one network with redundancy)

Governance Federal or State regulations City department, Commission

Ownership Private or Public

Decentralized Models in the Global North

Decentralized Models in the Global North

Single source, treated as needed, stored (often in a pressure tank in the basement)

Owned and maintained by the homeownerInitial local health department inspection Additional testing at homeowner’s initiativeExample… Household wells

EPA’s case for POU/POEEPA’s case for POU/POE

Public water supply consumers may not always possess the financial resources, technical ability, or physical space to own and operate custom-built treatment plants

Small drinking water treatment systems, such as Point-Of-Use and Point-Of-Entry (POU/POE) units, may be the best solution for providing safe drinking water to individual homes, businesses, apartment buildings, and even small towns

These small system alternatives can be used for not only treating some raw water problems, but they are excellent for treating finished water that may have degraded in distribution or storage or to ensure that susceptible consumers, such as the very young, very old, or immuno-compromised, receive safe drinking water

POU/POE ConcernsPOU/POE Concerns

The problem of monitoring treatment performance so that it is comparable to central treatment

POU devices only treat water at an individual tap (usually the kitchen faucet) and therefore raise the possibility of potential exposure at other faucets. Also, they do not treat contaminants introduced by the shower (breathing) and skin contact (bathing)

These devices are generally not affordable by large metropolitan water systems

POU devices are only considered acceptable for use as interim measures, such as a condition of obtaining a variance or exemption to avoid unreasonable risks to health before full compliance can be achieved

POE SolutionsPOE Solutions

The 1996 regulations required the POU/POE units to be owned, controlled, and maintained by the PWS or by a person

under contract with the PWS operator to ensure proper operation and maintenance compliance with the MCLs or treatment technique

equipped with mechanical warnings to ensure that customers are automatically notified of operational problems

Under this rule, POE devices are considered an acceptable means of compliance because POE can provide water that meets MCLs at all points in the home

Could each community in the Global South have a designated person who maintains the POU devices?

POU wins over Centralized Treatment when…

POU wins over Centralized Treatment when…

The distance between houses is large (order 1 km) then POU supplies are common

The centralized system is unreliable (low institutional capacity, poor infrastructure)

The cost of POU treatment is less than the cost of a centralized treatment facility (small communities)

POU only treats water for human consumption (with savings in capital, operation, and maintenance costs)

Opening QuestionOpening Question

You live in a small community that chlorinates a surface water with turbidities that range between 5 and occasionally 200 NTU

Give 2 reasons why a POU SSF might not be a good solution

What research would you like to conduct to determine how serious these problems are?

Water Quantity and Access for Health

Water Quantity and Access for Health

Service level Access measure Needs met

Level of health risk

No access (quantity collected often below 5 l/c/d)

More than 1000m or 30 minutes total collection time

Consumption cannot be assured Hygiene - not possible (unless practiced at source)

Very high

Basic access (average quantity unlikely to exceed 20 l/c/d)

Between 100 and 1000m or 5 to 20 minutes total collection time

Consumption - should be assured Hygiene - handwashing and basic food hygiene possible, laundry/bathing difficult to assure unless carried out at source

High

Intermediate access (average quantity about 50 l/c/d)

Water delivered through 1 tap on-plot or within 100m or 5 minutes total collection time

Consumption assured Hygiene - all basic personal and food hygiene assured; laundry and bathing should also be assured

Low

Optimal access (average quantity 100 l/c/d and above)

Water supplied through multiple taps continuously

Consumption - all needs met Hygiene - all needs should be met

Very low

Reactor Challenges for POUReactor Challenges for POU

Flow rate controlBatch vs. continuous flowQuantity of water to treatOperation and MaintenanceMonitoring (or the lack thereof)

is there any indication of whether the POU device is working?

Failure modes… HACCP

Water Safety PlanWater Safety Plan

Risk assessment to define potential health outcomes of water supply

System assessment to determine the ability of the water supply system to remove pathogens and achieve defined water quality targets (remember the chlorinator assignment?)

Process control using HACCP Process/system documentation for both steady

state and incident-based (e.g., failure or fault event) management

Hazard Analysis at Critical Control Points (HACCP)

Hazard Analysis at Critical Control Points (HACCP)

It is recommended that HACCP for household water collection, treatment and storage be applied in the context of a Water Safety Plan that addresses source water quality, water collection, water treatment, water storage and water use.

HACCP for Household Water Storage Vessels

HACCP for Household Water Storage Vessels

Hazard Vessel Type Vessel Integrity Vessel Sanitation

Critical Control Point(s)

Appropriate or not appropriate, based on design

Intact or not intact, based on visible damage (e.g., cracks, scratches), broken or missing parts (e.g., cap) and leaks

Sanitary or nor sanitary, based on frequency of cleaning and cleaning method

HACCP for Filtration/ChlorinationHACCP for Filtration/Chlorination

Type of Treatment

Source Water Hazards

Source Water Critical Control Point(s)

Treatment Hazards

Treatment Critical Control Points

Filtration methods

Contaminated or uncontaminated? Turbid?

Choose best available source, with low turbidity

Poor filtration and turbidity reduction

Observe (monitor) for adequate turbidity (cloudiness) reduction

Chlorination or mixed oxidants from electrolysis of brine (NaCl)

Contaminated or uncontaminated? Turbid? Chlorine-demanding solutes?

Choose best available source, with low turbidity and low chlorine demand

Poor chlorination due to inadequate dose and contact time

Observe (monitor) for chlorine residual (C) and for adequate contact time (T), i.e., adequate CT

HACCP for Boiling and SODISHACCP for Boiling and SODIS

Type of Treatment

Source Water Hazards

Source Water Critical Control Point(s)

Treatment Hazards

Treatment Critical Control Points

Heating to boiling with fuel

Contaminated or uncontaminated?

Choose best available source

Inadequate temperature achieved

Heat to a visible rolling boil

Solar Radiation in clear plastic bottles (heat + UV radiation or heat only)

Contaminated or uncontaminated?Turbid? UV-absorbing solutes?

Choose best available source, with low turbidity and low UV-absorbing solutes

Inadequate sunlight to achieve target temperature and UV dose

Target temperature sensor (thermometer or melting wax); elapsed exposure time (timer, clock, sun position, etc.); monitor/observe weather (sunny, part sun or cloudy)

Reflections…Reflections…

We need better solutions for Particle removalChemical removal

Existing designs are too expensive, don’t work well enough, or require advanced operator skills

We need easy to use and cheap monitoring devices Remove particles before disinfection (unless you are

using heat) Can we outperform PuR? We need better guidance for technology selection based

on turbidity (or other easily monitored parameters)

Two meanings!

Monitoring CapabilitiesMonitoring Capabilities

Chlorine disinfection – measure residualHach $0.27 to $1.25 per test

Too expensive for POU applicationsReasonable for community systems

Monitoring Capabilities: Coliform Monitoring Capabilities: Coliform

Current cost is several dollars per sample for membrane filtration (enumeration)

Absolutely prohibitive for POU monitoring Difficult for small communities MIT Design that matters is exploring cheaper methods of

measuring coliform concentrations Melted wax incubator More economical filtration apparatus

Coliform removal is still one of the best ways to evaluate filter performance (remember bacteria are hard to remove)

Testing for Coliform Bacteria:Presence/Absence Tests

Testing for Coliform Bacteria:Presence/Absence Tests

Colisure allows testing for coliform bacteria and/or E. coli in 24 - 28 hours.

The detection limit of ColiSure is 1 colony forming unit (CFU) of coliform bacteria or E. coli per 100 mL of medium.

If coliform bacteria are present, the medium changes color from yellow to a distinct red or magenta.

If E. coli are present, the medium will emit a bright blue fluorescence when subjected to a long wave (366 nm) ultraviolet (UV) light.

Testing for Coliform Bacteria: Membrane Filtration

Testing for Coliform Bacteria: Membrane Filtration

Membrane filter0.45 μm pores47 mm in diameter

Filter 100 mL of water to be tested through the membrane filter

Membrane FiltrationMembrane Filtration

Petri dish with sterile

absorbent nutrient pad

Add 2 mL of m-endo broth

(selective media)

Place membrane filter in the

petri dish on top of the

nutrient pad

Membrane Filtration:Incubation and ResultsMembrane Filtration:

Incubation and Results

Incubate for 24 hours at 35°C

Coliform bacteria grow into colonies with a green metallic sheen

Non-coliform bacteria may grow into red colonies

Coliform concentration is __________________

12

34

5 6

78

8 coliform/100 mL

Monitoring: TurbidityMonitoring: Turbidity

Hach portable Turbidimeter: $837.00Sechi disk (great for lakes…)SODIS technique

Turbidity MeasurementsTurbidity Measurements

90° detector

lamp

lens

sample cell

0° detector

LED

sample cell

10° detector

Turbidity Sensors (approximate turbidity measurement)

Cheap Turbidity MeasurementsCheap Turbidity Measurements

What is our cheap detector?What is the detector measuring?

How could you make a cheap method of measuring turbidity

eye

Refraction Transmission

1 10 100 1k 10kpeople

100k

1 10 100 1k 10k 100kLiters

person day

1 10 100 1kLiters

person day