Filter Media Treatment – Fall 2011

Po-Hsun Lin

Abstract:

The post sedimentation addition of polyaluminum chloride (PACl) was investigated as a means to enhance particle removal efficiency in rapid sand filtration. The process modification was evaluated in laboratory studies and at the Cornell Water Filtration Plant (CWFP). PACl was continuously metered into CWFP filter influent to increase concentrations by 0.06 to 4.2 mg/L (as aluminum) during the filter-to-waste stage of the filter operation cycle to accelerate filter ripening. Lower influent PACl concentrations ranging from 0.056 to 0.43 mg Al/L were also continuously applied during filtration. In comparison to a control filter that received no PACl addition, the ripening time required decreased with PACl dose, and the incremental improvement in particle removal during filtration increased with PACl dose. The addition of 0.056 mg Al/L of PACl (the lowest concentration tested) significantly reduced initial filter ripening time at the CWFP from 10 hours to 2.5 hours, and effluent turbidity in the test filter over the 77 hour filter run was lower than the control filter by an average of 17%. Incremental head loss increase caused by the PACl feed was dose dependent and was negligible for the lowest dosage tested.

Computational Fluid Dynamics Flocculation Tank Simulation, Fall 2009

Abstract:

Computational Fluid Dynamics (CFD) is a tool used by AguaClara to obtain a description of the flow through a portion of the plant where particle formation and growth (flocculation) occurs. This part of the plant is referred to as the flocculator. In this section, dirty (turbid) water flows through a series of baffles that enhances turbulent mixing. Essentially, for particles (flocs) to grow and eventually settle out in the sedimentation tank, they first must collide. By increasing the level of turbulence, the flocculator is increasing the collision potential of flocs. However, if there is too much turbulence, the flocs will break up and not settle out. By running CFD simulations of the flocculator, AguaClara can analyze the parameters important to flocculation and use the resulting data when making design decisions.

Computational Fluid Dynamics Flocculation Tank 3D Simulation, Spring 2009

Abstract:

The flocculation tank simulation team works on building a stable and reliable numerical model to simulate the flow inside the hydraulic flocculation tank, and providing well-studied guidelines for design, construction and operation of the flocculation tank.

Gravity powered hydraulic flocculators are used by AguaClara small-scale water treatment plants due to their low cost, inherent simplicity and robust operation. However, their inflexibility of energy input into the water relative to mechanical flocculators requires well studied design based on the understanding of the flow field and relevant performance parameters.

An appropriate CFD simulation can provide detailed numerical solutions for all the variables in the flow field, and by varying parameters such as tank geometry and flow conditions, we could obtain predictions of each of the flow variables and thus optimize the design towards lower cost and better performance.

Currently, our effort is focused on depicting an accurate energy dissipation map inside the flocculator, describing the size and the shape of the region where most of the energy is dissipated and the formation and collisions of flocs happen, thus providing basis for more efficient and economical design utilizing geometries that dissipate energy as uniformly as possible.

Chemical Dose Controller Retrofit Designs, Fall 2009

Abstract:

In some AguaClara plants, a surface foam develops at the end of rapid mix. The initial focus of the research was on the chemical conditions required for this surface foam to develop then the focus shifted to the fluid mechanics that make this occurrence possible and the simple retrofit designs that can ameliorate these conditions. In the initial experiments, different chemical conditions were tested for using a series of jar mixers and one-gallon tanks that modeled rapid mix. The first few trials tests ran a constant supply of clay with varying amounts of alum but these did not exhibit any form of surface foam formation. Subsequent trials included organic matter: humic acid, but these only produced large non persistent bubbles. It was not until a stronger surfactant, liquid soap, was added to the baffle spacing that a surface foam with strong persistent bubbles developed. From these experiments it was concluded that air entrainment along with a surfactant in the raw water are the main chemical factors behind surface foam formation.

Upon observing that waterfalls, like the one found in the LFOM, created the ideal fluid dynamic conditions for air entrainment; the second half of the research focused on retrofitting the LFOM at current AguaClara plants. The four designs that were suggested either used a submerged orifice, a vertical surface area or an inclined plane to decrease the velocity of the incoming water through the LFOM. In testing the viability of each design option the three limiting parameters of foam formation from water jets were recognized and documented.

Chemical Dose Controller Surface Foam, Summer 2009

Abstract:

In some AguaClara plants, a surface foam develops at the end of rapid mix. The initial focus of the research was on the chemical conditions required for this surface foam to develop then the focus shifted to the fluid mechanics that make this occurrence possible and the simple retrofit designs that can ameliorate these conditions. In the initial experiments, different chemical conditions were tested for using a series of jar mixers and one-gallon tanks that modeled rapid mix. The first few trials tests ran a constant supply of clay with varying amounts of alum but these did not exhibit any form of surface foam formation. Subsequent trials included organic matter: humic acid, but these only produced large non persistent bubbles. It was not until a stronger surfactant, liquid soap, was added to the baffle spacing that a surface foam with strong persistent bubbles developed. From these experiments it was concluded that air entrainment along with a surfactant in the raw water are the main chemical factors behind surface foam formation.

Upon observing that waterfalls, like the one found in the LFOM, created the ideal fluid dynamic conditions for air entrainment; the second half of the research focused on retrofitting the LFOM at current AguaClara plants. The four designs that were suggested either used a submerged orifice, a vertical surface area or an inclined plane to decrease the velocity of the incoming water through the LFOM. In testing the viability of each design option the three limiting parameters of foam formation from water jets were recognized and documented.

Turbidimeter, Spring 2011

Emily Clamp, Rohiverth Guarecuco, Julia Morris

Abstract:

The goal of the turbidity team was to create a low-cost turbidimeter that measures water turbidity within the range of 5 NTU to 250 NTU. Thus far, the team has brainstormed various turbidimeter designs and created several prototypes for simultaneously testing different LED display patterns. Many patterns have been assessed, including a dual-range LED display pattern for measuring a broad turbidity range. The dual-range LED pattern was tested using an experimental setup that allowed turbidity measurement of water that was constantly mixed with kaolin clay using a water pump. The team determined that only the fine pattern of the dual-range pattern was necessary, since the pattern alone could accurately measure turbidities from 5-200 NTU. This approach is based on resolution of the fine pattern within the turbidimeter as opposed to the use of contrast when using conventional Secchi disk patterns. The fine-resolution pattern was used to create a low-cost turbidimeter prototype equipped with an NTU scale based on the power-law equation derived from experimental results.

Tubidimeter, Summer 2011

Jennifer Gass, Maxwell Petersen, Heidi Rausch

Abstract:

The goals of this summer’s Turbidity team were to:

• Finish the testing that the previous team had been working on in order to design a cheap (under $20) Turbidimeter that can easily be transported to potential AguaClara facility locations.

• Find a relationship between depth and Turbidity that is within 50% accuracy for a specific disk design based on line thickness and spacing.

• Fabricate and calibrate 10 turbidimeters that will be ready for shipment by July 28, 2011.

Thus far, the team has managed to improve the design of the original Turbidimeter while lowering cost and increasing portability. All ten prototypes were built, calibrated, and sent to Honduras by the specified date. The only shortcoming was that due to size limitations the Turbidimeter could not measure below 15 NTU, however, this does allow for greater ease of use. The final Turbidimeter design is just over 60 cm in height and costs $4.02 to make.

Turbidimeter, Fall 2011

Julia Morris, Andrew Gorodetsky, Heidi Rausch

Abstract:

This report will cover all the work that has been done by Cornell’s AguaClara program on turbidimeters. Research first started on creating a new, low cost turbidimeter at Cornell in the Spring of 2011. Since then several different prototypes have been created and ten turbidimeters have been sent to Honduras for use by communities who are considering building an AguaClara plant. The reason that a low cost turbidimeter needs to be developed is so that communities who may be in need of water treatment facilities can test their water without incurring the high expense of other turbidimeters currently on the market. The most current complete turbidimeter prototype can read NTU values down to 15 NTU. The research discussed in this report details new turbidimeter designs with which it may be possible to read NTU values down to approximately seven NTU. The most promising design includes the use of a blue LED light and a large HDPE block, which is used for diffusing the light. However, this design will need to be tested more thoroughly for accuracy before it can be fabricated for use in the field. In the future if research continues to be done to try to create a turbidimeter that can read turbidity values below 5 NTU the length of the lowering rod may have to be made longer than the current prototype, which is only 60 cm long. Without adding length to the lowering rod current research suggests that it may be impossible to read the turbidity of any water with an NTU value lower than seven.

Structures, Spring 2011

Hyeong Yoon, Thomas Shouler, Joe Beaudette

Abstract:

Our main objective we wished to accomplish this semester was to create a means of automating the design of the columns and walls for the design tool. We began this work by analyzing the structural capabilities of the columns and walls for the Alauca plant.

There are three different load cases that guide the design of the columns and walls of the tank. The first case assumed that the tank walls were supported by the surrounding backfill. The second case assumed no support from this backfill. The third case analyzed the structural importance of the rubble which lies at the base of the sedimentation tank.

For our analysis of the walls, we modeled them as closely spaced columns. As seen in Figure 1, the vertical rebar that runs through these walls will add flexural support. By modeling the walls as columns we accounted for this vertical rebar. Modeling the walls as a combination of individual columns also allowed us to use the same tools and procedures that we used for the analysis of the columns. We set the moment at initial cracking as our first failure moment. We determined this moment by using the the Transformed Moment of Inertia method. We also wanted to know the moment that would render the walls and columns could experience before ultimate failure. We believed that this value would be of importance for future analyses which would incorporate earthquake conditions. For this ultimate failure analysis, we used the Column Interaction Diagram. This method plots the area of all axial load and moment cases that a column would be able to support safely. All methods are explained in detail in the report.

Structural Design of AguaClara Plants, Summer 2011

Lily Siu

Abstract:

Our main objective we wish to accomplish this summer is to analyze the reinforcement configuration and structural strength of the sedimentation and flocculation tank walls. In the previous semester, the structural design team analyzed the structural capabilities of the columns and walls for the Alauca plant using various assumptions and load cases. The previous team analyzed the walls as closely spaced concrete columns. By modeling the walls as columns the flexural support provided by the horizontal reinforcement is unaccounted for, but it allowed for the use of the same tools and procedures that is used for beam analysis. We seek to attempt to validate the previous team’s calculations as well as suggesting methods to analyze the horizontal reinforcement in order to reduce over-designing. This report is meant to augment the Spring 2011 report.

Sedimentation Tank Hydraulics, Summer 2011

Elana Liskovich, Mahina Wang, Jill Freeman, Yiwen Ng

Abstract:

A floc blanket is a dense fluidized blanket of flocs that helps to reduce effluent turbidity in the sedimentation tank by trapping other flocs. The geometry of the sedimentation tank is crucial in determining the extent of floc resuspension by the jet and hence floc blanket formation. To improve tank bottom geometry, nine experiments were conducted, each testing a different tank bottom geometry. The experiments were run in a 1/2 inch wide tank to model a thin slice of the full scale sedimentation tank. The geometry that resulted in the least sludge accumulation and therefore best floc resuspension was two 60 degree inserts leading to a semicircular trench 10 cm in diameter. We also provided initial designs and calculations for a floc weir to maintain the height of the floc blanket. A preliminary experiment was also conducted to evaluate the feasibility of our initial floc weir design.

Sedimentation Tank Hydraulics, Fall 2011

Jill Freeman, Mahina Wang, Saied Khan, Matthew Hurst

Abstract:

A floc blanket is a dense, fluidized bed of particles that forms in the sedimentation tank and helps to reduce effluent turbidity by trapping small flocs and reduces clean water waste through less frequent draining of the sedimentation tank. Floc resuspension is necessary for floc blanket formation so that flocs are recirculated through the tank instead of settling on the tank bottom as sludge. Research was conducted to examine mechanisms for floc resuspension. Parameters important for floc resuspension include energy of the jet stream on its upward flow path, position of the jet as it interacts with solids, and hydrodynamic pressure of the jet compared to hydrostatic pressure of the returning solids. Several geometries were tested with red dye and fully built floc blankets to observe the jet path and velocity profile around the bottom geometry. Best results are achieved through geometries that preserve jet momentum, especially through splitting the jet flow, and geometries that maintain a high jet velocity when contacting solids. Later, quantitative measurements were taken to determine floc blanket performance for various bottom geometries.

Residuals Management, Fall 2011

Patrick Farnham

Abstract:

The Residuals Management sub-team is solving the problem of solids disposal in AguaClara treatment plants. Currently, settled solids from the sedimentation and entrance tanks are drained and routed directly onto the nearby landscape. The newly created stacked rapid sand filtration system will produce backwash water in need of disposal, and spikes of highly turbid influent water bypass the plant by being discharged down the surrounding slope. The research goal is to determine inexpensive and responsible disposal methods for these outflows as well as for precipitate matter removed from the chemical stock tanks. Flow rates and concentrations of all residual flows have been estimated with the help of AguaClara engineers in Honduras, and designs have been created for pipe outlet protection structures which should reduce the erosive power of AguaClara residual flows. The team goal is to identify promising methods and eventually code them into the AguaClara design tool for use in the future and also for possible use in retrofitting current plants.

Foam Filtration, Summer 2012

Michelle Gostic, Sarah Levine, Leah Meyerholtz

Abstract

A reticulated polyurethane foam filtration system combined with coagulant dosing has proven capable of providing clean, low-turbidity water; however the necessary addition of coagulant combined with a non-conventional cleaning method makes the foam filtration system inefficient and not ideal to be used on a municipal scale. For this reason, research on foam filtration methods have been geared towards engineering a portable and effective filtration unit to provide clean water to devastated areas in emergency situations. There is a great demand for such technology as proven by the 1994 Rwanda Crisis. In this case, 85-90% of deaths in refugee camps were caused by diarrhea [Toole and Waldman; Doocy and Burnham]. Providing a low-cost, sustainable clean water source in emergency situations is the best way to prevent waterborne diseases and dehydration. The foam filtration team has worked towards designing an apparatus that will provide at least 15 liters of water per day per person, based on the UNHCR’s (United Nations High Commissioner for Refugees) recommendations for refugee situations [UNHCR]. The foam filter is small enough that it can be placed in the back of a pick up truck when needed and hooked up to the car’s engine to harvest electricity needed to power the filtration process. The current AguaClara foam filtration system utilizes a roughing and finishing filter to remove solid particles from influent water. Turbid water enters through the linear flow orifice meter (LFOM), which maintains a linear relationship between the water level and the flow rate of water through the system. Coagulant enters the LFOM in regulated doses to ensure consistent mixing of coagulant with influent water. The water flows through a 30 inch deep roughing filter, consisting of 30 ppi (pores per inch) foam, and then through a 15 inch deep finishing filter, consisting of 90 ppi foam. Before effluent water leaves the system, it is dosed with chlorine and exits the supercritical flow tube. Previous tests showed that head loss from the foam filters was negligible, but recent experimentation with an open system has proven otherwise. Due to the head loss and effluent turbidity standards, the foam filters need to be cleaned. A “plunger” method of cleaning has proven to be an efficient and easy way to clean the foam filters. A long pole with a porous disc attached to one end is used to compress the foam and release solid particles trapped in the 1 filters. This method is similar to squeezing out a dirty kitchen sponge. During the cleaning process, the foam must remain submerged in water to prevent the entrainment of air bubbles that would hinder filter performance. After plunging, the dirty water and released sediment flow out of exit valves located downstream of each filter. At the 2012 National Sustainable Design Expo in Washington D.C. the AguaClara team showed that the foam filter system is effective in removing particles from water at relatively high turbidities; however, data from these tests does not give concrete evidence of the filter’s performance in terms of realistic applications. At the expo, the turbidimeter was not properly calibrated, thus the data collected is not 100% accurate. Additionally, the filters were cleaned every 4 hours. Although this proved that the filters could return to their initial performance level after being cleaned via the plunger method, cleaning every 4 hours may be either too frequent or infrequent to maintain a high volume of treated water with acceptable effluent turbidity, in an emergency situation it is essential to maximize the volume of usable water while still maintaining quality standards. Effluent at the expo was dosed with clay and recycled so that the effluent became the influent. There is a possibility that coagulant built up in the filter columns, due to the recycle, improved performance beyond what it would have been in a more realistic experiment. As a new and relatively unexplored technology, there is much research to be done in the realm of foam filtration. Our current and future research will focus on providing extensive data as to the performance of the foam filtration system at different turbidities as well as identifying more efficient cleaning cycles. The World Health Organization included significantly reducing the number of “people without sustainable access to safe drinking-water and sanitation” in its list of 15 Millennium Development Goals [WHO]. Refining the foam filtration system has the potential to contribute to achieving this goal in addition to providing clean water to those who need it the most.

Foam Filtration, Spring 2012

Katie Edwards, Walker Grimshaw, Bradshaw Irish, Kelly McBride, Nadia Shebaro

Abstract

The Foam Filtration team developed a two-stage emergency water lter. The lter was presented at the 2012 National Sustainable Design Expo in Washington, D.C. as part of the P3 competition for sustainability. While at the expo, rough data was collected on the performance of the pilot scale lter as a proof of concept. The competition also involved submitting a grant proposal highlighting all past research on foam ltration and presenting a plan for future research and eventual implementation.

Laminar Tube Flocculator, Fall 2012

Patience Ruijia Li

Abstract

According to the predictive occulation model proposed by Swetland et. al., 2012, large ocs do not signicantly contribute to turbidity removal  only small colloids can collide eectively and aggregate to a size that will be removed by sedimentation. Based on the hypothesis that large ocs are useless, a oc breakup procedure was devised. Results obtained using a coiled tube occulator and occulation residual turbidity analyzer (FReTA) shows that higher turbidity removal was achieved after breaking the ocs, comparing to results using the same method but without oc breakup. Therefore breaking ocs at regular intervals to maintain continuous growth will promote better performance of occulation. This research nding provided a good reference for future hydraulic occulator design.

Low Flow Flocculator, Spring 2012

Ryan Anthony, Elyssa Dixon, Zac Edwards

Abstract

Open ow occulators, like those currently used in AguaClara plants, increase in cost for lower ow rates (less than 5 L s ). AguaClara must be able to meet the need of a wide range of community sizes, so scaling the design for low ow plants is crucial. This report specically analyzes three dierent scenarios that use obstructions within a pipe to remove the geometric constraints that cause the errors found in the current designs. Two scenarios include placing semi-circular baes (similar to those in open ow occulators) within the pipe and maintaining the spacing between and above the baes or maintaining the area between and above the baf- es. The last scenario involves placing balls on a string through the center of a pipe. A table is provided within this report comparing critical values for these occulators for a 3 L/s plant. An initial comparison of the three options shows that the scenario that maintains area above and between baes is the most eective with materials. However, discrepancies with calculations in the approach using balls as obstructions ultimately yields the results inconclusive.

Sedimentation Tank Hydraulics, Spring 2012

Jill Freeman, Mahina Wang, Matthew Hurst, Saied Khan, Yiwen Ng

Abstract:

A floc blanket is a dense, fluidized bed of particles that forms in the sedimentation tank. It helps to reduce effluent turbidity by trapping small flocs and reduces clean water waste through less frequent draining of the sedimentation tank. Floc resuspension is necessary for floc blanket formation so that flocs are recirculated through the tank instead of settling on the tank bottom as sludge. Research was conducted to examine the effectiveness of the retrofitted Marcala sedimentation tank. At high influent turbidities, a steady floc blanket was obtained, but performance was slightly compromised when the influent turbidity was lowered to simulate Marcala conditions during the dry season. A floc blanket visibly formed with an influent turbidity of 5 NTU after about 1 week but “seeding” the tank with coagulated flocs will minimize floc blanket formation time. Images were also acquired for hindered sedimentation velocities of 0.6 mm/s, 1.2 mm/s, and 1.6 mm/s and analyzed with a floc-water interface program using a region of interest to better understand hydraulic processes within a floc blanket. Complete settling curves from this data confirmed wall effects significantly affect settling velocity. A floc hopper proved to be effective at controlling the height of the floc blanket when the accumulated flocs were drained at an adequately high flow rate. A lower alum dose of about 39 mg/L for an influent turbidity of 100NTU resulted in a less sticky sludge that could be more easily drained from the hopper.

Sedimentation Tank Hydraulics, Summer 2012

Danielle Feng, Jill Freeman, Cari Gandy

Abstract:

In the sedimentation tank of an AguaClara water treatment plant, water flows through the inlet manifold with vertical diffusers that channel the water into the bottom of the tank as a line source. As water exits the vertical diffusers, a semi circular half pipe jet reverser directs the water upward to resuspend flocs to form a floc blanket, or a dense, fluidized bed of particles, in the sedimentation tank below the plate settlers. A floc blanket increases the particle removal efficiency of the sedimentation tank by capturing smaller flocs that would otherwise escape through lamellar sedimentation. A floc blanket also leads to less clean water waste because without a floc blanket, sludge builds up at the bottom of the tank and will require constant draining. While current plant designs use a 0.5” radius jet reverser and centered jet placement, other jet reverser sizes and jet placements were explored to increase floc resuspension and floc blanket stability. A 1.5” radius reverser with asymmetric jet placement was found to be the optimal design for floc resuspension. Floc blanket stability in relation to coagulent dose was also explored, and optimal alum doses for several influent turbidities were determined.

Sedimentation Tank Hydraulics, Fall 2012

Frances Ciolino, Hongyi Guo, Ethan Yen

Abstract:

The sedimentation tank hydraulics team this semester focused on optimizing the floc hopper. Our main goal was to learn more about the floc hopper geometry and reasons for floc blanket failure. We started the semester by looking at the vertical sedimentation velocity. This velocity is controlled by the flow into the tank and affects how fast the particles settle. If this velocity is too high or too low, the sedimentation tank will not form a proper floc blanket.

We also looked into how the plan view area of the floc hopper would effect the floc blanket formation and performance. Changing the size of the floc hopper effects how much of the area allows for up-flow of the water and how much captures the flocs. In our current experiments we are trying different sizes at different wasting rates. We are looking for a size and rate that keeps the plant running efficiently, meaning the least amount of water wasted while keeping the water leaving the plant clean.

One of the ideas we are working on is continuous wasting of the flocs. The idea behind this is to allow for constant removal of flocs instead of collecting the flocs and then manually opening a valve to let them leave the tank. When the wasting rate is optimal, the flocs will be allowed to compact before they are removed so that the least amount of water is lost in the process. Our last experiment looked at finding the proper wasting rate where the rate of particles flowing into the floc hopper is the same as the rate at which the particles are being removed.