CHNG3807 Report - Revised Ed_002

111Equation Chapter 1 Section 1CHNG3807 Products and Value ChainsSchool of Chemical and Biomolecular Engineering

Faculty of EngineeringUniversity of Sydney

Chemical Product Design:Water

Name SID

Date Submitted: 2nd September 2009

1

Contents0.0 Executive Summary.................................................................................................................3

1.0 Background Information.........................................................................................................4

2.0 Needs..........................................................................................................................................42.1 Specifications..................................................................................................................................5

3.0 Ideas...........................................................................................................................................53.1 Ceramic Filters................................................................................................................................53.2 Disinfection.....................................................................................................................................63.3 Slow sand filtration.........................................................................................................................63.4 Idea Screening.................................................................................................................................6

4.0 Selection....................................................................................................................................74.1 Cost Analysis..................................................................................................................................74.2 Effectiveness...................................................................................................................................74.3 Availability......................................................................................................................................74.4 Less Subjective Criteria..................................................................................................................74.1 Mechanism of a Slow Sand Filter...................................................................................................8

5.0 Design........................................................................................................................................95.1 Sizing of the sand filter...................................................................................................................95.2 Bacterial Kinetics..........................................................................................................................105.3 Filtration Mechanics.....................................................................................................................115.4 Pressure drop through the sand filter............................................................................................125.5 Cleaning and maintenance............................................................................................................125.6 Safety Assessment – In progress...................................................................................................135.7 Prototype Design...........................................................................................................................145.8 Cost Summary of the water production unit (incomplete)............................................................15

6.0 Design Summary....................................................................................................................16

7.0 Table of Equations.................................................................................................................18

8.0 Appendix.................................................................................................................................198.1 Calculations...................................................................................................................................198.2 Figures...........................................................................................................................................208.3 Tables............................................................................................................................................22

9.0 References...............................................................................................................................23

2

0.0 Executive SummaryThe aim of this investigation was to design a product for a third world country that would effectively solve their drinking water treatment problems. Our chosen region was a small community residing near the river Nile just outside the central city of Khartoum, the capital of Sudan. The main issue with the water supply to this community was the presence of coliform bacteria in the water that reaches levels of about 2000 counts/dl during the rainy season. This is very much higher than the Environmental Protection Agency (EPA) standards for clean drinking water. (Source)

So to keep it realistic, our group decided to design a slow sand filtration system for a small group of 4 or 5 households with a maximum of 10 people requiring at least 200 litres of water every day. Following the steps of a successful chemical product design process, we first brainstormed ideas from a range of sources and narrowed them to a few which we then analysed using different selection criteria.

3

1.0 Background InformationThe provision of drinking water of adequate quality and quantity remains a major public health issue in many African countries. According to the World Health Organisation, 39 000 Sudanese people died due to improper water, sanitation and hygiene (2008). One of the principle areas affected by this problem is Wadramli, a small village located on the peripheries of Khartoum City. The village is characterised by crowding, poor housing, contaminated water and sanitation related problems.

Wadramli has a population of approximately 5 000 people, and is situated on the bank of the river Nile, 70 km north of Khartoum (the precise location is shown in Figure 1 (Musa, Shears, Kafi, & Elsabag, 1999)). Currently, untreated water is obtained by being pumped from the Nile into an overhead storage tank and distributed through taps outside homes.

Figure 1 – Location of Wadramli near Khartoum on the River Nile (Musa, et al., 1999)

The whole of Sudan is very poor. Although the GDP has grown substantially in the last decade, a United Nations Development Programme study ranked the country 141st out of a total of 177 for the percent of the country living in poverty; 65 – 75 per cent of people in North Sudan, where Wadramli is located, live below the poverty line of $1 per day and rural areas are particularly hard-hit (United Nations Development Programme Sudan, 2009).

2.0 NeedsThe primary problem associated with the water quality is the presence of micro organisms that make the water detrimental to the health. The current water quality was assessed via conducting a literature review. The main problem with the water out of the Nile in our selected area is the bacteria coliforms in the river; a study by Dirar (1985) detailing the bacteriological content of the Nile near Khartoum revealed a count of faecal coliforms, in the worst months, of nearly 5000 per 100 mL.

4

A different study found that over the course of one year the cases of diarrhoea within the Khartoum province ranged from around 2 500 per month up to 8 500 per month in the monsoon. Faecal coliform counts within the water were found to have reached as high as 2000 counts per 100 millilitres (Musa, et al., 1999).

Another study by Attar, Gawad, Khairy, & Sebaie (1982) of the parasitological content of water in the Upper Nile Delta in Egypt revealed a significant amount of protozoan cysts within the water, with roundworm found in 15 per cent of the samples tested and threadworm in 10 per cent.

A thesis by Abdalrahim (2007) detailed the water quality of the River Nile, at Khartoum, before it splits into the Blue and White Nile. The results are summarized in Table 6 in the Appendix which shows that there are no major chemical (in terms of metal ions and other indicators) concerns in the River. The following matrix of needs (Table 1) summarises the needs of the local Sudanese village.

Essential Useful Desirable– Enough water for 20 people;– Removal of bacteria, viruses, protozoa as well as solids;– Affordable.

– Easy to use;– Safe to operate (i.e. in terms of chemicals);– Low dependence on replacement parts/chemicals;– Reliability.

– No manual labour required to operate;– Taste;– Portable;–Environmentally sustainable;– No external power required.

Table 1 - Categorising Needs

2.1 SpecificationsThe United States’ Environmental Protection Agency’s (EPA) has the standard that the ideal level is a coliform count 0 count per 100 mL – the public health goal – and a “maximum contaminant level goal” of a count of 5 per 100 mL (Environmental Protection Agency, 2009b). A study by Dirar (1985) detailing the bacteriological content of the Nile near Khartoum revealed a count, in the worst months, of nearly 5000 per 100 mL.

We have a target population of 20 people. According to De Zuane (1997) eight litres of water per person per day for both drinking and cooking is required, this value can be scaled up for our population unit. For twenty people we will need to deliver around 200 litres per day of filtered water, allowing for over-engineering.

3.0 Ideas3.1 Ceramic FiltersCeramic filters function by obstructing the passage of particles larger than a water molecule. In most cases, they are treated with silver in the colloidal form to hamper the growth of mould, algae and bacteria. The generic design of these filters has a receptacle placed beneath the filter to collect the purified water.

Ceramic materials are brittle in nature and therefore allow the formation of hairline cracks that are not clearly visible. This will allow larger particles to penetrate the filter and thus defeat the purpose of the device.

3.2 Disinfection

3.2.1 Solar Trough Treatment

5

This method utilizes sunlight as a source for disinfecting water. The simplest form is the solar box, which can be used for water pasteurization. It consists of a cardboard box, painted black on the inside, and a clear plastic lid. A covered pot, polished black, is placed inside the box where it needs to remain at 65oC for a few minutes. This method has a capacity to pasteurise 3.5 litres in three hours on a very sunny day (Rolla, 1998). Another application of this technology is the “parabolic trough reactor,” which concentrates parallel rays of the active ultra violet part of the solar spectrum by a factor of 30 to 50. Flow rates of 500 to 3000 litres per hour can be achieved in this device (Goslich, Dillert, & Bahnemann, 1997).

3.2.2 Chlorination (Benchmark Technology)Chlorination is chosen as the benchmark technology as it is an effective way to kill and inhibit the growth of microorganisms in water. It is the most common disinfectant in Australia and therefore is used in many Australian cities. The availability and cost of chlorine raises questions on the viability of this method in our target area, Sudan. On the upper hand, chlorination has numerous impressive benefits such as, effective at low doses, controls odour and taste, provides residual protection and thus avoids recontamination, easy to apply, control and monitor. There are some limitations to chlorine in that it produces unfavourable by-products, it is not effective against cryptosporidium and it needs to be transported and stored in a chemical form (Science Applications International Corporation, 1994).

3.3 Slow sand filtrationFiltration processes are used to reduce turbidity and microorganism levels in the water supply, and has proven to be quite effective in removing pathogens and viruses. Slow sand filtration provides a more biological approach by passing dirty water through a sand bed. The complete process involves multiple stages mainly classified under transport, attachment and purification mechanisms (Huisman & Wood, 1974).

Biological filters are easy to operate, daily routine checks for the water temperature, turbidity and the filter head loss are required to maintain constant functionality. Depending on the raw water quality, a slow sand filter can run without being cleaned for about twenty to ninety days. Slow sand filters can be cleaned by manually removing the top few inches of the sand which can either be thrown away, or washed and kept for later use. A technical limitation is the fact that it requires at least one or two days after the cleaning procedure for the sand system to redevelop

Water systems with relatively high turbidity do not work well with slow sand filters; in ideal conditions, a turbidity value of less than 20 NTU is recommended. A cost study conducted in 1992 showed that a slow sand filtration system with a capacity of 190 000 litres per day would represent a construction cost of $210 000 and an annual operational and maintenance cost of $7 000 (Science Applications International Corporation, 1994).

3.4 Idea ScreeningOrganisation of ideas is crucial to lead to a more quantitative and engineering based analysis of feasible technologies. The table below enlists the more relevant ideas in what is called an ideas screening matrix (Cussler & Moggridge, 2001). It is important to realise that this is merely a qualitative approach to our problem with objective criteria deduced from the already established list of needs in the section above.

6

Criterion Ceramic Filters Sun UV Rapid sand SSF ChlorinationEssentialCapacity + + + + +Cost - + - + +Effectiveness - - + + +DesirableUser friendly - + + + +Safe + + + + -Maintenance + + - + -Reliable + - - + +

Table 2: Ideas screening matrix

4.0 SelectionThe next step in product design was to select the best possible idea based on a certain set of selection criteria, which has both subjective and objective components, employing a more engineering based approach to choose the finest option.

4.1 Cost AnalysisThe chosen water treatment system should be able to produce the required 200 litres of water per day with an E. coli count as close to the EPA standards as possible (Schiller & Droste, 1982). It should be cheap, reliable, safe and easy to maintain. Due to the poor economic conditions of the Wadramli people, the limiting factor when choosing a technology is economic not technical. The government normally covers capital costs but in most cases, it is the high maintenance costs that lead to the failure of a water treatment system (Huisman & Wood, 1974). Ceramic filters are sophisticated and expensive technologies so can be easily ruled out based on this investigation.

4.2 EffectivenessThe second most suitable factor deciding the feasibility of a system is its effectiveness as a treatment option. Chlorination is a proven method of water treatment that is used in developed countries to purify water. Considering the target volume of water is 200 litres per day which translates into 95 litres of chlorine required per year (US EPA, 2006), and a price of around $3 to 4 a litre of household bleach, this means an annual cost of nearly $400, which is not an ideal situation in a third world country.

4.3 AvailabilityConsidering chlorination, it must be compared with other forms of disinfection in terms of effectiveness. For water to be cleaned by solar pasteurisation it must reach and stay at a high temperature for a given period of time. Sources differ between the required temperature (and time) required to safely treat water, however it is recommended by the World Health Organisation (2008) to bring water to the boil. Considering a 10 × 10 m2 solar box situated at the equator, it would require approximately 15 hours of continuous heating to achieve the required 200 litres per day (Seider, Seader, & Lewin, 2004). This fails to meet the criterion of easy availability for the Sudanese villagers who need it most during the warm summer days.

4.4 Less Subjective CriteriaAnother relevant factor worth mentioning is a technology’s scientific maturity (Cussler & Moggridge, 2001). Theoretical designs of many technologies are available in journals and textbooks but the use of these treatment systems in practice is a very important issue, which needs to be

7

addressed before singling out the best possible design. For example, water treatment by slow sand filtration is often considered inefficient and old fashioned but, contrary to popular belief, this system of water purification has been in continuous use since the nineteenth century (Huisman & Wood, 1974).

Chlorination was introduced as a major advancement in water treatment, which would potentially remove all harmful material from water to make it suitable for drinking purposes. Initially it was considered as a whole technology to be used on it’s own for water disinfection rather than using it as an effective stage in a combination treatment system. For example, chlorination could be employed as a water disinfecting technique when the primary technique is malfunctioning or is shut down for maintenance. The main reason for chlorination not being a treatment system on its own is the fact that the water entering a chlorination unit needs to be heavily pre-treated for effective disinfection. When measuring effectiveness, there are a few essential things a filtration system must be able to do:

1. The chemical forms of dissolved organic and inorganic material are broken down or precipitated so as to be easily removed in the next stage

2. Maximum amount of living matter must be removed by settlement supported by a chemical treatment if necessary

3. Enough residence time is provided to kill all the microorganisms in storage (Huisman & Wood, 1974).

Slow sand filtration incorporates many of these functions in a single unit operation and is therefore our suggested water treatment system. The selection matrix below quantifies our findings mentioned above to show a clearer picture of why slow sand filtration is our preferred technique to solve the drinking water problems of the people of Wadramli.

Criteria Weight SSF

Chlorine* Solar

Cost 0.3 8 5 3Efficiency 0.3 6 5 5Maintenance 0.2 6 5 4Scientific maturity

0.2 8 5 4

Total 1 7 5 4Table 3: Selection Matrix in reference to Chlorination (benchmark tech.)

4.1 Mechanism of a Slow Sand FilterSlow sand filters consist of a layer of bacteria called the schmutzdecke that is cultivated within a bed of filter medium. The schmutzdecke acts as a bio filter removing harmful bacteria such as E. Coli and salmonella from contaminated water. The bio-layer utilises these pathogens as its substrate, consuming them as their food source, and effectively removing them from the water.

The bio-layer is created by continually passing contaminated water through the sand bed. Initially, sand has a slight negative charge that attracts positively charged species such as iron cations. Certain areas in the sand become oversaturated with positively charged species and attract negatively charged organic colloidal particles such as bacteria, which would have otherwise been repelled by the sand and passed through, become trapped and begin to form the schmutzdecke layer.

These microorganisms then release a slimy substance known as zoogloea that forms a film attracting more organic matter which helps complete the schmutzdecke. Viruses, protozoa and pathogens that

8

pass through the bio-layer provide either energy or a building material for the trapped bacterium. The amount of substrate present in the contaminated water is limited, this bacterial growth is coupled by a dying period.

In order for these bacteria to survive they require a constant water head of approximately 1 meter and the turbidity below 20 NTU (Public Health Management Guide). As a result, slow sand filters tend to be a continuous process with an initial pre-screening filter that lowers turbidity i.e. a nylon mesh. The schmutzdecke layer removes organic matter and some other impurities while the nylon mesh lowers the waters turbidity. The sand bed provides a breeding site for the microorganisms but also acts as a mineral oxidation zone that further reduces the waters turbidity. Any remaining large particles are trapped within the sand either by the sands small pores or through electrostatic attraction. Finally a gravel layer at the bottom helps support the sand and provides minimal obstruction to the water as it exits the bed and into the pipe. The water that emerges from this pipe is safe and ready for consumption.

9

5.0 Design and ManufacturingFrom the outcome of the selection process a slow sand filtration system was chosen as the preferred system. When designing a slow sand filter the main design considerations are the size and depth of the filter bed, the filtration rate and the bacteria removal by the schmutzdecke. The main design parameters have been evaluated and compared to values extracted from other research studies (see Table 4). (DOES THE TABLE NEED TO BE HERE/OR REFERED TO FROM THE APPENDIX)

Source Depth (m)

Sand size (mm)

Sand bed depth (m)

Gravel depth (m)

Filtration rates (m/h)

Pressure head (m)

Downing, et al. (2002) 0.46 No data 1 0.2 – 0.3 No data No dataVan der Hoek, Hofman, & Graveland (2000)

0.8 – 1.2 0.2 - 1 0.8 – 1.2 No data 0.2 – 0.5 No data

Van der Hoek, Hofman, Bonne, Nederlof, & Vrouwenvelder (2000)

No data No data No data No data 0.25 No data

NDWC (2002) No data 0.15 – 0.35 > 0.7 No data 0.15 0.7 – 1.5Cleasby & Logsdon (1999)

No data 0.3 0.9 0.45 - 0.6 0.07 – 0.12 No data

Eighmy, Collins, Malley, Royce, & Morgan (1993)

0.5 - 1 0.15 – 0.3 No data No data 0.1 – 0.2 2

Barrett, Bryck, Collins, Janonis, & Logsdon (1991)

≥ 1 0.2 – 0.3 0.8 – 0.9 0.3 – 0.5 0.1 – 0.2 1

Table 4 – Compilation of various design parameters by notable Bio-sand studies (Page et. al, 2006)

5.1 Mass Balance and Sizing

5.1.1 Surface areaGiven that the velocity of water for safe slow sand filtration is around 0. 1 m hr -1 to 0.5 m hr-1 (Page et al, 2006) and the water flow rate through the system is defined as 200 L day-1(2.315×10-6m3s-1).Therefore the surface area was be adjusted to meet this requirement. (REFER TO EQUATIONS)

Mass in = Mass Out (Mass Balance)200 L/day = 200 L/day

212\* MERGEFORMAT (.)

From this and using a diameter of 0.3 m, we are given a surface area of 0.071 m2 and a velocity of 0.12 m hr-1. (DOESNT MAKE SENSE/WE CAN USE AN EQUATION HERE)

5.1.2 Filter DepthThe removal of biomass by the schmutzdecke is the initial mechanism in slow sand filtration and is active 0.3 to 0.4 metres in the top layer of sand. If this is a minimum depth for bacterial removal only, a reasonable total depth of 1 metre can be assumed based on the values in (See Table 4). The removal efficiency depends on fluid velocity and the grain size used (Huisman & Wood, 1974).

10

5.2 Bacterial KineticsAssuming that the substrate growth within the filter is minimal, a simplified kinetic model can be fitted to simulate the dynamics with the aid of Matlab’s Simulink package. This is comprised of a system of three differential equations detailed below.

Firstly, the Monod Equation was modelled using the bacterial concentration as the limiting growth factor:

313\* MERGEFORMAT (.)Secondly, two differential equations were modelled to highlight the interaction between the substrate and biomass (bacteria and protozoa);

414\* MERGEFORMAT (.)

515\* MERGEFORMAT (.)

11

5.2.1 Simulink ResultsThe kinetic model used in section 5.2 shows a distinct correlation between the amount of bacteria and the development of the schmutzdecke. It shows that as the predator population increases within the schmutzdecke, the amount of bacteria begins to decrease, eventually reaching a steady state value of 55 counts per 100 mL after 400 hours. The predator population continues to grow and eventually reaches a steady state level of 20 000 counts per 100 mL at around 900 hours.

Figure 2 – Bacterial concentration (predator – red and prey – blue) as a function of time

The bacteria reach a steady state at around 400 hours, therefore the deduced start up time is approximately 16 to 17 days. Temperature has a directly proportional relationship with the growth of bacteria and protozoa. This can be exploited to help decrease the start up time by using higher temperatures making schmutzdecke grows faster and is more efficient at removing bacteria.

A possible problem is that of a clean water spike. This was simulated using the same model in Simulink. It was found that it takes around 500 hours for the predator levels to reach zero counts per 100 mL (see Figure 8). To combat the issue of schmutzdecke decay, it is recommended that the water above the filter be seeded by a small amount of contaminated water daily. This can simply be done by adding a small amount of animal manure to water and mixing it to the part of the filter above the schmutzdecke (Palmateer, et al., 1999)

5.3 Filtration MechanicsThe second portion of the system involves the sedimentation of larger objects (usually greater than 10 µm) being filtered by the small pores formed by the packed sand. If sand particles of 0.15mm are used, the smallest pores are approximately 20 µm in diameter. For one cubic meter of sand the gross surface area for sedimentation is given by formula (1.5), assuming the area of sand particles is one continuous surface (Huisman & Wood, 1974).

616\* MERGEFORMAT (.)12

Predator

Prey

According to Huisman & Wood (1974), even allowing for surfaces not facing upwards and other resistances to sedimentation this area is easily in excess of 1000 m3. With this area and a flow rate of 1.1 m hr-1 this gives a surface loading of 0.1 × 10-3 m hr-1. Using Stokes formula for settling velocity, the maximum size of particles that can be removed is determined.

717\* MERGEFORMAT (.)

The settling velocity versus the surface loading can then be presented as an inequality to determine the size of particles that are deposited.

818\* MERGEFORMAT (.)Hence sedimentation is capable of only removing particles in the order of 3 to 4 µm.

5.4 Pressure drop across the sand bed The Darcy’s laminar pressure drop equation,

919\* MERGEFORMAT (.)Where:

10110\* MERGEFORMAT (.)The equation above assumes the ideal situation in which the sand in the device is clean. In practise, the sand is clogged with impurities that have been collected over time. Bourget, Gatellier, & Hermin (1988) found that during the end of a slow sand filtration cycle, up to one-third of the pores in the sand become congested, reducing the porosity to around 0.2. As a worst case scenario, this represents a head loss of about 0.7 m. For this reason, Page, Wakelin, van Leeuwen, & Dillon (2006)recommend a pressure head of at least 1.5 m.

13

5.5 Cleaning and maintenanceSlow sand filters need to be cleaned in order to maintain an acceptable flow rate. As the flow rate decreases, the water is of a higher quality due to a greater residence time. The relationship between the efficiency of a filter versus flow rate can be seen in (Figure 3). When the flow rate is unacceptably slow, the filter needs to be cleaned; the pre-filter should be cleaned on a regular basis to ensure that it is not clogged with large particles which can inhibit the flow rate.

Figure 3 - Typical Performance of a SSF (Manz, 2009)

The slow sand filter itself needs to be cleaned of dirt particles accumulated in the top layers of the sand. This involves filling the filter with water, blocking the exit, and carefully swirling the water without making contact with the sand (Fewster & Mol, 2009). The movement of the water loosens the accumulated dirt which come into suspension and can be removed along with the dirty water. The process is repeated a number of times, until the water is visibly cleaner. Upon restarting the filter, the flow rate should increase significantly (Fewster & Mol, 2009).

The final cleaning stage is for when terminal head loss develops. In this case, the top one to two centimetres of the sand – which includes some of the bio-layer – needs to be scraped off and removed. The sand can be washed and later replaced (Logsdon, Kohne, Abel, & LaBonde, 2002).

5.6 Safety Assessment – In progressAn overall safety assessment was done to evaluate how safe our product was for public use and the risks associated with its use. Educating all stakeholders in our venture is critical to the safety of everyone.

Certain scenarios relating to a slow sand filters misuse or harsh external conditions have the ability to hamper its performance and safety. Firstly, the schmutzdecke has to be free from any sort of physical, biological or chemical stress, as this affects the quality of the performance of the filter. The film of bacteria should not be broken as it would impair its performance greatly. In addition to that, oxygen supply is critical to the function of this device as the bio-layer present survive by respiration, utilizing nutrient sources like dead pathogens and biodegradable contaminants. Suppose the oxygen levels fall to nil, anaerobic decomposition processes drive the schmutzdecke to produce hydrogen sulphide, ammonia and other metal complexes. As a result, the drinking quality of water is hampered.

14

Temperature is another key factor that controls the safety of this process and when it’s low the microbes lose their activity levels significantly. Therefore the harmful organisms are consumed at a slower rate, having a higher likelihood of passing the biological filter and into the clean water. This significantly decreases the removal efficiency of bacteria, putting the drinker at risk of contracting water borne diseases. For example, the E. coli removal factor is reduced from 1000 to about 2 at a temperature of around 2°C.(http://www.biosandfilter.org/biosandfilter/index.php/item/320 ).

The filtration rate strongly determines the effluent quality as unexpected changes disturb the bacterial equilibrium. Therefore abrupt changes in raw water quality should be avoided and continuous operation is recommended. The rate at which the quality of water deteriorates is not very significant with filtration rates therefore it is still potable to a certain extent. In summary, the filter would dysfunction due to all the above scenarios and produce turbid and pathogen rich water.

Failure of mechanical supports due to various reasons could cause the device to collapse and fragment causing total failure. The PVC piping has to be sealed carefully at points where they join the tanks and bends. They should be checked for leaks first. The three drums used in our design have to be airtight and hole-proof so that no water leaks out and there is no contact with foreign media.

Risks Hazard Prevention Monitoring Corrective Procedure

Bad raw water quality

Dangerous Micro-organisms not removed

Check raw water quality

If the raw water quality exceeds 20NTU (Public Health Management Guide)

Add pre-treatment mechanisms

Incompatible flow rates

Low water quality output

Monitor the Flowrate

Flowrate should not exceed 0.5 m/h (Page e. al, 2006)

Change Flowrate

Undeveloped Schmutzdecke

Low water quality

Avoid overloading it and let the Schmutzdecke develop

Counts of microorganisms coming out

Allow the Schmutzdecke

Sand layer lacking oxygen

Low water quality

Avoid the growth of algae and ensure a continuous flow through the system

Odours in the water

Degrading material in the sand layer

Allow for mixing before the schmutzdecke

5.7 Prototype DesignThe prototype we designed was a scale model of the design postulated within this report: the diameter of the rim was the same, but the height of gravel and sand was different. We included a baffle delivery systems capable of automatically diffusing water into the sand filter – the main benefit of this is that the bio-layer would not be disturbed by the slow trickle.

It was built using a bucket, plywood, PVC pipe and silicon glue. Boring drill bits were used to create holes which were then filed out for snugness and sealed water-tight with silicon. Refer to Figure 4 for a diagram

15

.

The tap (the outlet with the arrow in Figure 4) rises to just above the sand level so it can be assured the schmutzdecke does not dry out any time when the filter is not in operation. The results achieved with the scale were good with dyed water coming out clear – however, this is not representative of bacterial removal as it would take several weeks of flow with contaminate for a bio-layer to develop.

5.8 Cost Summary of the water production unit (incomplete)Material Quantity Cost(Unit price) Cost (AUD)Sand 136 kg $ 48/tonne 6.50Gravel 85 kg $0.33/kg 28Piping Pre-filter 1 1Plastic Tap 1 2Aeration Tank 1Cylinder(holds the filter) 1Holding Drum 1Valves 2The operating cost of the Bio sand filter is negligible as nothing gets used up in the process. There are no kinetic features that would need replacement and repair. The only form of maintenance would come in the form of replacing of the pre-filter. Once source indicates that a typical sand filtration system requires $ 1-2 for a period of 3 years.

(http://pcv_nathan_beckett.tripod.com/mylifeinranquittehaiti/id4.html)

Therefore, the cost of the potable water is the cost of the device itself and an American made filter is around $USD 15. The labour component required to assemble the device is insignificant as household labour can be used.

16

D ≈ 30 cm

Sand

Gravel

Figure 4 – Prototype schematics (left) and baffle system (right) which fits inside the rim of the housing for the slow sand filter

5.8.1 Materials of Construction Sand Gravel Pre-Filter Piping(PVC) Aeration Tank Holding Drum Sand Filter Cylinder Plastic tap 2 valves

The concrete, sand and gravel required for this device would be readily available near the surrounds of the target area.

6.0 Design Summary

From and the design calculations carried out above the final recommended design for the water purification system is as follows. The system consists of a pipe connected to the town’s main water tank flowing to the inlet of the sand filter. The pipes flow rate is regulated by a gate valve to maintain a constant supernatant water height of 1.5 m. The water passes through a pre-filter made ideally of a sheet of nylon mesh, to reduce the initial turbidity of the supernatant water.

The water then enters the sand filter tank which is 0.3 m in diameter, containing 0.3 m of supporting gravel and the bottom and 1 m of sand, combined with the height of supernatant water giving a total tank height of 2.6 m. With these dimensions the water flows through the sand at a velocity of 0.12 m/hr giving a total residence time that is the time it takes for a molecule of water to pass through the filtration system is close to 22 hours. The clean water then passes from the sand filter into a large holding tank ready for use. Shown below (Figure 5) is the final design for the slow sand filtration system.

17

Figure 5. PFD of final slow sand filter design

18

7.0 ConclusionSudan has a serious issue when it comes to supplying clean drinking water to its people. Due to the impoverished conditions, ordinary means of filtration and purification are too costly. When the principles of chemical process design were applied to the problem, Slow Sand Filtration emerged as the optimum choice over other options including; Ceramic filters, Solar Trough Treatment, and Chlorination.

The kinetic model of the schmutzdecke showed approximately 99% removal of bacterial coliforms. However in application it is estimated that around 80-90% of bacterial coliforms will be removed in the shmutzdecke and up to another 10-20% will be removed in the top 30-40 cm of sand. (reference)

It is necessary for the proposed slow sand filter to undergo periodic maintenance. The pre filter needs to be checked and cleaned regularly, and the upper layer of the schmutzdecke may need to be cleaned by the process called ‘wet harrowing’. These cleaning steps need to be implemented when a decrease in flow rate is observed. If terminal head loss develops however a few centimetres of the sand may need to be removed and replaced.Because the Slow Sand filter can be maintained and operated by the people who use it, many of the maintenance costs can be avoided. The estimated capital cost of the slow sand filter is (?) which equates to a running cost per person of $(?) day-1. This is well below the average daily income of $0.93.

The final design of the filter stands at a height of 2.8m and has a diameter of 0.3m.The filtered water then passes to a holding tank of capacity 2000L. This is a precaution incase the filter needs to be decommissioned for any period of time and ensures the people continue to have clean water for up to 10 days.

Bacterial Coliform pre filtration 5000 counts mL-1

Bacterial Coliform after filtration 55 counts mL-1

Fluid Velocity 0.12 m hr-1

Height of Sand 1mHeight of Water Head 1.5m

Residence Time 22 hr-1

19

7.0 Table of Equations Table Use First used in

µ Specific growth rate (1.5)µmax Maximum specific growth rate (1.5)

S Bacterial concentration (g/L) (1.5)

KS Maximum specific growth rate of bacteria (1.5)

Q Volumetric flow rate (L/ day) (1.6)

V Filter volume (L) (1.6)Xi Inlet biomass concentration (g/L) (1.6)Kd Death rate constant (day-1) (1.6)Si Inlet bacterial concentration (g/L) (1.7)Y Yield of biomass on bacterial (g biomass / g bacteria) (1.7)Pr concentration of predator organisms (1.8)

kg max pr Maximum specific growth rate of predator organisms (1.8)Sc Concentration of prey organisms (1.8)kss Half saturation coefficient for prey (1.8)kd pr Death rate of predator organisms (1.8)

kg max s Maximum specific growth rate of prey organisms (1.9)Nc Concentration of prey food (TOC, nitrates, et cetera) (1.9)kds Death rate of prey organisms (1.9)Ypr Growth yield coefficient for the predator (g predator / g prey biomass) (1.9)ε Porosity of the medium, 0.38 (1.10)ds Diameter of the particle, 0.025 mm (1.10)

∆P/p Ratio of densities of suspended matter and water (less that 0.01) (1.12)U Fluid velocity through the filter (m s-1) (1.14)H Height of sand (m) (1.14)

∆P Pressure drop (m) (1.14)K Permeability of sand (1.14)T Temperature of the medium, 25oC (1.15)φ Sphericity of the medium, assumed to be 0.95 (1.15)

20

8.0 Appendix8.1 Calculations

11111\* MERGEFORMAT (.)

Giving a velocity of:12112\*

MERGEFORMAT (.)

From Equation (1.5), where:

13113\* MERGEFORMAT (.)And from Equation (1.9)

14114\* MERGEFORMAT (.)Substituting into the Darcy equation (1.8)

15115\* MERGEFORMAT (.)

21

8.2 Figures

Figure 6 – Grain Size Analysis on Sand Particles (Huisman & Wood, 1974)

Figure 7 – Coliform Counts in the Nile over 12 months (Abdalrahim, 2007)

22

Figure 8 - Modelling of schmutzdecke clean spike

23

8.3 Tables

Sand filter with pump

Slow sand filter Boiling/Pasteurisation UV treatment

Activated carbon filter

Reverse Osmosis Membrane filter Ceramic filter

Chlorination Mixed oxidant gases systems

(MOGGOD)

Ozone treatment Magic

Fluidised bed Chemical flocculants

Dew – new source Ship water from cleaner source

Pump water from clean source

Parabolic trough reactor

Hot water coils Tow an iceberg

Nylon mesh Well Waste water treatment/recycling

Electrolysis

Solar box Flow through unit Solar still Seed flocculationSilver/Titanium Genetically

modified anti-bacteria bacteria

Aerobic reactor Anaerobic reactor

Sell bottled water (EVIANS)

Add alcohol Acid/base purification (CaSO4 insoluble salt)

Desalinate from another source

Research new chemical treatment

Centrifuge Drink juice Clean up the Nile

Carbon Nanotubes Radiation CFC’s from scrap Fridges

Table 5 – Ideas to Solve the Problem

Analysis Blue Nile White Nile Nile Canadian StandardTemperature (oC) 20 20 20 ≤15

pH 7.5 8.2 7.9 6.5 – 8.5Conductivity (µs/cm) 240 190 210 1000

AOX (µg/L) 32 11 9 30TOC (µ g/L) 5.9 9 7.4 15

Cd (µg/L) 0.008 0.002 0.024 5Pb (µg/L) 0.5 <0.1 0.2 10Cr (µg/L) 0.6 6.6 3.9 5

Nitrate (mg/L) < 0.002 0.011 0.032 3.2Table 6 – Summary of observed water parameters compared to MAC (Maximum Allowable

Concentration). Note: TOC (Total Organic Carbon) and AOX (Absorbable Organic Halides) (Abdalrahim, 2007)

Condition Symbol Value UnitsFilter volume V 0.1 m3

Flow rate Q 0.00833 m3 hr-1

Maximum specific growth rate µmax 0.042 hr-1

Initial Bacterial concentration Si 5 counts/100 mL

24

Initial Predator concentration Xi 0.001 counts/100 mLMax. specific growth rate of Bacteria 0.175 hr-1

Death rate constant Kd 0.0016 hr-1

Yield of Predator on Bacteria Y 0.5 mg biomass/mg bacteriaTable 7 – Conditions of the Kinetic Model

9.0 ReferencesAbdalrahim, O. (2007). Effect of Khartoum city for water quality of the River Nile. Linköping

University, Linköping.Attar, E. L., Gawad, A. A., Khairy, A. E. M., & Sebaie, O. E. (1982). The Sanitary Condition of

Rural Drinking Water in a Nile Delta Village. I. Parasitological Assessment of 'Zir' Stored and Direct Tap Water The Journal of Hygiene, 88(1), 57-61.

Barrett, J. M., Bryck, J., Collins, M. R., Janonis, B. A., & Logsdon, G. S. (1991). Manual of design for slow sand filtration: American Water Works Association Research Foundation.

Bourget, L., Gatellier, C., & Hermin, M. N. (1988). Microbial well-clogging and the circulation of water. Revue des Science de L'eau, 1(1), 3-21.

Cleasby, J. L., & Logsdon, G. S. (1999). Section 8.74: Granular bed and pre-coat filtration Water quality and treatment - A handbook of community water supplies. Texas: McGraw-Hill

Cussler, E. L., & Moggridge, G. D. (2001). Chemical Product Design. New York: Cambridge University Press.

De Zuane, J. (1997). Handbook of Drinking Water Quality (2nd ed. ed., pp. 575). New York: Van Nostrand.

Dirar, H. A. (1985). Coliform bacterial counts in the Nile water at Khartoum. Environment International, 71, 571-576.

Downing, J. B., Bracco, E., Green, F. B., Ku, A. Y., Lundquist, T. J., Zubieta, I. X., et al. (2002). Low Cost Reclamation using Advanced Integrated Wastewater Pond System Technology and Reverse Osmosis. Water Science and Technology, 45(1), 117-125.

Eighmy, T. T., Collins, M. R., Malley, J. P., Royce, J., & Morgan, D. (1993). Biologically enhanced slow sand filtration for removal of natural organic matter: AWWA Research Foundation.

Fewster, E., & Mol, A. (2009). Biosandfilter.org - Flow Rates Retrieved 01-09-09, from http://www.biosandfilter.org/biosandfilter/index.php/item/317

Goslich, R., Dillert, R., & Bahnemann, D. (1997). Solar Water Treatment: Principles and Reactors. Water Science and Technology, 34(4), 137-148.

Huisman, L., & Wood, W. (1974). Slow Sand Filtration. Geneva: World Health Organisation.Logsdon, G. S., Kohne, R., Abel, S., & LaBonde, S. (2002). Slow Sand Filtration for Small Water

Systems. J. Environ. Eng. Sci., 1, 339-348.Manz, D. H. (2009). BSF Guidance Manual 2: Maintaining or Cleaning the Concrete BioSand Water

Filte Retrieved 01-09-09, from http://www.manzwaterinfo.ca/documents/guidance/BSF%20Maint%20and%20Cleaning%20Jan%202009.pdf

Musa, H. A., Shears, P., Kafi, S., & Elsabag, S. K. (1999). Water quality and public health in northern Sudan: a study of rural and peri-urban communities. Journal of Applied Microbiology, 87, 676-682.

NDWC (2002). Technical brief – a national drinking water clearinghouse fact, slow sand filtration. National Drinking Water Clearinghouse.

Page, D., Wakelin, J., van Leeuwen, J., & Dillon, P. (2006). Review of Biofiltration Processes Relevant to Water Reclamation via Aquifiers. Adelaide: University of South Australia.

Palmateer, G., Manz, D., Jurkovic, A., McInnis, R., Unger, S., Kwan, K. K., et al. (1999). Toxicant and Parasite Challenge of Manz Intermittent Slow Sand Filter. Environmental Toxicology, 14, 217-225.

25

Rolla, T. (1998). Sun and Water: An Overview of Solar Water Treatment Devices. Journal of Environmental Health, 60(10).

Schiller, E. J., & Droste, R. L. (Eds.). (1982). Water supply and sanitation in developing countries. Ann Arbor: Ann Arbor Science.

Science Applications International Corporation (1994). Small Community Water and Wastewater treatment Retrieved 01-09-09, from http://books.google.com.au/books?id=El6tsJZfvewC&lpg=PP1&pg=PP1#v=onepage&q=&f=false

Seider, W., Seader, J., & Lewin, D. (2004). Product and process design principles : synthesis, analysis, and evaluation. New York: Wiley.

United Nations Development Programme Sudan (2009). UNDP Sudan | Achieving the MDGs and Reducing Human Poverty, 01-09-09, from http://www.sd.undp.org/focus_poverty_reduction.htm

Van der Hoek, J., Hofman, J., Bonne, P., Nederlof, M., & Vrouwenvelder, H. (2000). RO treatment: selection of a pretreatment scheme based on fouling characteristics and operating conditions based on environmental impact. Desalination, 127(1), 89-101.

Van der Hoek, J., Hofman, J., & Graveland, A. (2000). Benefits of ozone-activated carbon filtration in integrated treatment processes, including membrane systems. Aqua- Journal of Water Supply: Research and Technology, 49(6), 341-357.

World Health Organisation (2008). Safer Water, Better Health: Costs, benefits and sustainability of interventions to protect and promote health. Geneva: World Health Organisation.

26