- Abstract
- Introduction
- Literature Review and Previous Work
- Timeline of UASB
- Spring 2019 Goals
- Lab Experiments
- Conclusion: Future Work
- EPA Funding Assurance
For the past two years, the AguaClara Upflow Anaerobic Sludge Blanket Team, also known has UASB, has been in the works of designing and fabricating a pilot-scale Upflow Anaerboic Sludge Blanket reactor, a system which treats wastewater via anaerobic digestion, producing biogas in the process. UASB's design is completely gravity-powered and is a low cost solution to treating wastewater in rural communities that cannot afford typical wastewater treatment plants.
This semester, the UASB Team divided into the two sister teams of UASB Research (UASB-R) and UASB Design (UASB-D). UASB-R is focused on benchtop testing of a model UASB, researching ways to increase the reactor's efficiency, and test out designs in lab before being implemented by UASB Design.
When communities do not have access to wastewater treatment, they are forced to dump untreated wastewater directly to the environment. It eventually ends up in our rivers, streams, lakes, and oceans, and underground water sources tapped for well water (Yahner and Taylor, 2018). Wastewater contains organic and fecal matter, which can have harmful effects on the environment. Nutrients, such as nitrogen and phosphorus, can invade water sources causing algae blooms, and heavy metals and other inorganic material can also pollute bodies of water (Gao, 2017).
Untreated wastewater also is a significant public health risk, since wastewater can carry hundreds of different pathogens that can make a human very ill. Pathogens may be transmitted by direct contact with sewage, by eating food or drinking water contaminated with sewage, or through contact with human, animal, or insect carriers. Typhoid fever, cholera and especially diarrhea are common waterborne diseases that are often found in areas with improper wastewater treatment.
It is difficult for many communities to find technology that can offer a reasonable, inexpensive way of treating wastewater. The United States treats its wastewater with technology that requires large retention times and takes up significant amounts of land. American wastewater treatment plants also require a large energy input because they are usually designed for densely populated urban areas, and are therefore expensive to set up and run.
The UASB reactor is designed to reduce the challenges of setting up wastewater treatment systems in developing countries. An UASB reactor works by pumping wastewater into the bottom of a large tank filled with colonies of methane-producing, anaerobic bacteria, that form a structure called a sludge blanket. As the wastewater works its way up through the sludge blanket, the bacteria break down the organic and fecal matter and produce methane as a by-product. This has two benefits: one, this removes harmful particles in wastewater, and two, the methane automatically bubbles up to the top of the tank, where it can be collected and be burned as an energy source or heating. The water will then exit out the effluent pipe, where it can then go through a secondary treatment such as chlorination. AguaClara’s UASB design, unlike conventional UASB reactors, is completely gravity-powered and therefore requires zero electrical energy input. However it requires an elevation difference to drive the flow of the wastewater through the UASB tank. This will be taken care of as long as the reactor is placed downhill from the wastewater source (e.g. a village).
A depiction of the design is shown below:
Fig. 1a and 1b: Diagram of a traditional UASB Reactor (front and side views)
A recent, updated model in OnShape has also been created by UASB-Design.
Tamil Nadu Agricultural University did a study in 2013 on an Upflow Anaerobic Sludge Blanket Reactor that was designed to handle 8,800 liters per day of influent wastewater where they found that the most effective volume of the reactor was 8.84 m-3 and that the optimum organic loading rate or the amount of water per unit of water was observed to have a chemical oxygen demand (COD) of 2.67 kgm-3 per day. This is the amount of oxygen needed to break down the wastewater's organic contents (so a high COD value is an indicator of a high amount of organic pollution). When the reactor operated at 3 days hydraulic retention time, or the length of time water has to sit in the reactor, the efficiency of the reactor was 70%. The reactor was also designed to collect gas that could be utilized for energy. This study done at a papaya fruit processing factory, where the reactor cost only Rs. 80,000 ($1160), showed that the gas used as energy had a payback period of 3 to 4 years, which is a fast payback period compared to solar panels. It also proved that the UASB provides onsite solutions to waste management problems (Jeyapandiyan, 2014).
From the article “Anearobic Treatment Using UASB Reactors: Engineering and Operational Aspects” (Jules B. Van Lier et al.) it's apparent that the UASB reactor consists of 4 functional units: Primary Clarifier to remove the non biodegradable solids, Biological Reactor (secondary treatment) to remove organic compounds, Secondary Clarifier to treat effluent in the settler zone, and Sludge Digester to stabilize and improve dewatering characteristics. The article also discusses the importance of a pre-treatment and post treatment units, and gave examples of how developing countries like India do not have the support from academic institutions for researching this process when they compared it to Brazil, a country that has invested in fundamental and applied research in the subject matter. What Brazil has done, resulted in a large variety of STP (Sewage Treatment Plant) in which the anaerobic technology is combined with adequate post treatment that offering cost-effective solution. Therefore, to point out the importance of the pre-treatment process, Lier mentioned that “without an efficient pretreatment treatment, the design, operation and maintenance of UASB reactors become an impossible job”. The site visit of full-scale STP’s revealed that there are many factors that are as important as a good design: [most importantly,] the operation and maintenance of the treatment plants (Lier, 2010).
The UASB team was formed in the Summer of 2016. At that point, the objective of the team was to design and implement a functional lab-scale UASB reactor to treat synthetic blackwater. Blackwater is wastewater from toilets; synthetic blackwater is a mixture of proteins, lipids and starches mixed together (usually with a few other chemicals as well; recipes for synthetic wastewater vary) In the summer of 2016, the UASB team constructed four small-scale UASB reactors and found that those small-scale reactors could successfully treat synthetic wastewater. The team also found that increased wastewater concentration and higher residence times resulted in higher biogas removal and increased Chemical Oxygen Demand (COD) removal.
In the fall of 2016, the UASB team underwent several changes. First, the team altered the design of the small-scale UASB reactors by implementing shorter and narrower influent lines. The team also changed the synthetic wastewater recipe. More specifically, the team substituted glucose with insoluble carbon compounds. The team also researched retention time in the reactors with fluoride tracer tests, and found a HRT of 3.22 hours in one of the reactors (close to the target of 4 hours).
During the spring of 2017, the team worked on assessing the efficiency of two modifications to the UASB reactor: plate settlers and the submerged gas collection lid. After conducting granule settling tests, the team was unable to conclude whether or not plate settlers significantly improve solid retention time. The team also conducted a Submerged Gas Capture Lid Test and concluded that a check valve would be useful to allow for continuous collection into a gas tank without loss of biogas.
In the summer of 2017, the UASB team continued with similar research and testing to that of the Spring 2017 semester. This time, the team determined that large, bulky plate settlers were not required for a full scale reactor. Instead, a smaller settling apparatus, such as a sloped exit weir, can achieve similar sludge retention time(SRT).
During the fall of 2017, the UASB team made several strides. Firstly, the team settled on the critical design assumptions for the fabrication of a UASB reactor at the Ithaca Area Wastewater Treatment Facility (IAWWTF). The critical design assumptions were as follows: a reactor diameter of 3 feet, a reactor height of 7 feet, a 4 hour hydraulic residence time, and a flow rate of .036 L/s. The team also designed a biogas capture lid. In this system, as biogas is produced in the reactor, it displaces water out from under the lid. Gas is then stored there until it is manually removed.
The team also wrote Python calculations in Jupyter notebook (available on GitHub) to estimate biogas production based on COD input.
Next, the team decided to add plate settlers back to the effluent tube. Plate settlers are intended to prevent dislodged granules from escaping the reactor. The fall 2017 team also worked on designing a sludge weir for the UASB reactor.
UASB continued efforts to improve the design of the pilot-scale UASB reactor in Spring 2018. A major area of focus was the influent flow system. After extensive research, the team chose top influent flow - where the wastewater flows down the inlet pipe until it reaches the bottom of tank - in order to prevent clogs. Next, the team decided to incorporate pulse flow into the reactor to achieve the desired flow of .03 L/s. The team proposed two methods to produce pulse flow: a tipping bucket design and a siphon.
UASB also worked on biogas production. The team wrote code to estimate the COD concentration of influent and the temperature of the reactor, as well as the biogas production rate based on the flow rate through the reactor. The team decided on a gas bag system to store biogas as it is flexible, easy to connect to the reactor, affordable, and easily transported. The team also wrote code to determine the required volume and dimensions for the lid of the biogas system. The team also looked into a fats, oils and greases (FOG) removal system. The conclusion was that manual skimming of FOG could always be used as a backup option.
Next, the team developed the design for the pilot UASB reactor's sludge sampling and removal system. For the sludge weir, the team chose a tube with a 6 inch diameter jutting out of the reactor at the predicted height of the sludge blanket at a downwards angle with two valves.
Lastly, the team wrote code to calculate the optimal size of the pilot UASB reactor's tube settler, the number of plates required, and the overall height of the settling arm.
The summer team continued work on the hydraulic design of the UASB system. The major design project was creating the tipping bucket system to deliver wastewater in pulses. The system consists of a pivoting bucket which fills with wastewater before tipping and delivering the wastewater in a large pulse at a high flow rate. After brainstorming the idea, the team designed and tested models to determine optimal design and flow rates for the system. The team also continued developing code for the entire system, and made a Computer Aided Design (CAD) model of the UASB reactor using Fusion 360.
During the fall of 2018, the team tested a benchtop UASB model in the lab. Because actual sludge granules were not readily available, and the team thought that using them in the early testing stages would be wasteful, the team decided to use tapioca to model the sludge blanket. Tapioca has similar material properties as sludge, inexpensive and readily available made it a suitable alternate material. Tapioca was used to identify preferential paths in the UASB reactor. Preferential pathways are the uneven distribution of wastewater flow throughout the reactor; the presence of these make the UASB reactor less efficient.
The benchtop model was a scaled down version of a real UASB reactor, made using Polyvinyl Chloride (PVC). The influent was set on top perpendicular to the bottom plate and a hole was drilled at the top of the model connected with a pipe as the effluent. The team used ProCoDa software to pump the water through the influent pipe. They also added red dye to the initial pulse of water to manually to look for preferential paths. It was suggested to use a second pump for inputting the dye in the future for more accurate testing.
The result of the testing shows that the layer of tapioca did not lift entirely off the bottom of the reactor, but instead it stretched as influent “wastewater” traveled through it. The red dye in the influent revealed the formation of preferential pathways through the tapioca layer. The team then came up with the idea of spreading the inlet flow by drilling four holes on the inlet tube near the bottom to decrease the preferential paths, but it was not completed by the end of 2018.
Afterwards, UASB tried to use a photometer to more accurately depict how the water was flowing through the sludge blanket by measuring whether or not the red dye concentration in the effluent tube was changing as expected. However, the tapioca was not a suitable material because it was too cloudy for the photometer to measure the red dye concentration.
The team then tried to use chia seeds instead of tapioca, which resulted in a worse scenario; the chia seeds was too buoyant in water and prone to flowing out of the effluent. Ultimately the team used glass marbles as the sludge blanket due to the time constrains. As predicted, this material was too heavy for the influent water to lift them and it was back to the drawing board.
The goal of this semester was to redefine the inlet system to allow even distribution of influent water and biogas capture. The model inlet design used during the fall was the single inlet tube whose goal was to deliver wastewater to the bottom of the reactor. But it was not spreading out the water evenly throughout the model sludge blanket leading to an inefficient reactor, thus less water purification. That is due to some sludge granules not getting into contact with the wastewater, and therefore not degrading the organic matter in the wastewater. Another significant problem with the previous inlet design was that the inlet tube was inserted directly above the UASB reactor, but this is where the biogas capturing system must be.
With the aforementioned problems, and the fact that UASB's old model was lost, creating a new model-scale inlet was a necessity. The team first sketched a few inlet designs with four openings to better distribute water flow. An idea of utilizing a manifold was shut down over concerns of clogging. After finalizing the designs, UASB-R got the parts to construct a new model and started fabrication. Shown in Figure 3 is Design #1 excluding the attachment with the four openings.
Fig. 3: Beginnings of the first inlet design of 2019
As one can see, the inlet tube enters on the right side of the tank, perpendicular to the tank wall. UASB-R also thought of inserting the inlet tube at a lower height to see if the water velocity and/or the headloss differed depending on where the inlet was inserted. The two designs are shown in Figure 4.
Fig. 4: Two competing designs; sketch does not include effluent design or the biogas capture system.
Below is a formal, more detailed list of the goals of UASB-R this semester:
- Construct and test the four inlet tube attachment using sludge granules.
- Test to see whether pulse flow vs continuous is flow more efficient using sludge granules
- Originally, UASB was under the impression pulse flow would increase efficiency in the UASB reactor by increasing upflow velocity, but UASB Design has been pointing toward continuous flow. Thus, we mainly research with continuous flow.
- Keep with UASB Design for any bench testing - specifically working on a design to evenly distribute wastewater flow through the inlet when the tipping bucket releases the wastewater.
UASB-R created an experiment that examined the exit velocity of the water from the effluent tube. Inlet systems that come in at two different heights of the benchtop model as shown in Figure 4 were designed and created.
UASB-R tested the first design (the sketch on the right in Fig. 4 and shown in Fig. 5) using only tap water. Several leaks were caused by the elbows of the model, thus they had to be replaced. The team also discovered leaks between the drilled hole and the 0.5" inlet pipe. To fix this, the team drilled a larger hole and attached seal tape to the threaded fitting to prevent leaks, as shown in Fig. 6.
Fig. 5: The old faulty connection
Fig. 6: the fixed connection
The UASB-R was started experimenting with old sludge. Note: While sludge can stay alive in a suspended state in a fridge for up to one year, with only a small amount of sludge left, the team planned to get more sludge in early March. Unfortunately, due to AguaClara's contact for sludge granules at the Aanheuser-Busch Plant in Syracuse, NY retiring, both Prof. Ruth Richardson of the CEE department and UASB-R did not have a contact so plans to get more sludge granules were stalled. Thankfully, the small amount was enough to finish this semester's testing.
UASB-R started pumping water through the UASB inlet at different flow rates and calculating the exit velocity. However, after consulting with a research advisor, it became apparent that an inlet tube at a lower height would not be beneficial as the different heights would not make a significant difference to the efficiency of the reactor. Also, the lower inlet would make the system harder to clean and fix if anything were to break. Thus, the remainder of the testing and some data initially collected on the exit flow velocity was scrapped, but this initial testing helped the team learn more about ProCoDa - the software used to run the pump - and the design in Fig. 6 was able to be reused in the next task of the semester.
The four inlet system is shown below. The goal was less about fabricating an model that would perfectly emulate the inlet system and more about fabricating a simple design quickly, with materials already in the lab, that would provide four openings for the water to flow into the reactor tank. The 1/2" pvc pipe split into two pipes using a wye connector, and those two pipes each had their own wye connector to split the pipe into two, creating four 1/2" openings. This was done so that the team could test the hypothesis: that a greater number of inlets would reduce preferential pathways through the sludge blanket. There is would be no use using this more complicated design if it were to give the same results as the initial one-inlet design from last semester.
.JPG?raw=true)
Fig. 7 - The inlet system with four openings
The reactor tank (made from a 4 and a 1/2" pvc pipe welded to a sheet of pvc at the bottom) was filled with tapioca and a bit of water, as shown in Fig. 8. A small amount of 20.0 g/ml solution of red dye was added to the tube that connects the pump and inlet, and water was pumped through the reactor, using ProCoDa at a flow rate of 1.0 ml/s. This one done four times. After each test, the tapioca was taken out of the reactor and rinsed with water to remove any leftover dye that would affect the team's results. Pictures and video recordings were taken to compare the results of each tests with each other and the tapioca tests with the one-inlet design from last semester.
.JPG?raw=true)
Fig. 8 - The tapioca-filled UASB reacotor, pre-testing.
[You guys should have a materials list to recreate both apparatuses that you have created, or at least make a note to refer to your sister team's report for materials. From the rubric "Clearly explain your apparatus and experiments. Provide enough detail for future team members and people unconnected to AguaClara to reproduce your experiments and apparatus" FIXED CJMS]
As water moved through the reactor, the initial burst of dye spread throughout the lower part of the reactor, however the dyed water did not reach the sludge blanket uniformly for the first 10-20s (Fig. 9 and 10), which could lead to delay in wastewater treatment on a bigger scale.
.JPG?raw=true)
Fig. 9 - Front view of the reactor
.JPG?raw=true)
Fig. 10 - Back view of the reactor
But we found that preferential pathways (which would look like a long, mostly vertical stripe of white tapioca) were not indeed forming. The water eventually reached each part of the tapioca and not selectively choosing certain pathways, as shown in Fig. 11 after running the water for about 40 seconds during one of the tests.
Fig. 11 - Reactor after the dye had spread further
However, UASB-R knew this wasn't a perfect solution, despite the significant improvements. During a meeting with Professor Weber-Shirk, Professor Richardson and the rest of the wastewater teams, there was a debate on how to check if the sludge blanket itself (and not just the tapioca model) is being properly fluidized in an UASB reactor. However, this idea would have to researched further before it can be implemented.
Video recordings of some of the testing can be found here and here.
The team continued the idea of using the tipping bucket from 2018 (i.e. where the wastewater flows out of a tipping bucket into the inlet after a certain amount of wastewater is collected to create a pulsed flow). Based on this idea, UASB-R fabricated an inlet system using four pipes, where the tipping bucket pours the wastewater, as shown in Figure 12.
Based on the sketch in Figure 12, the red section represents a wastewater tank that holds wastewater before entering the tipping bucket, and the black box located on the bottom of the drawing is the four pipe outlet system that deliver the waste water evenly to the sludge blanket reactor. The system can be called the "flow distribution system".
Fig. 12 - Sketch of the tipping bucket system. Inlet tubes not included
A clear PVC sheet was used to fabricate the bottom part of the system, so the team could observe what happens inside the tank. The upper part of the system was fabricated using a 6 quarts plastic gallon container with the diameter of 8", and the four pipes outlet system was fabricated using half inch thick plastic sheet. The clear PVC sheet was cut into squares with a perimeter of 3" and with a total of five sides. The four sides were for the wall, and the other side was used to enclose the bottom of the box. The four pieces of cut plastic then were glued together using PVC glue. The four 1/2 inch diameter holes were drilled into the bottom and then four of 1/2 inches pipes were attach into them.
Fig. 12.a - the flow distribution system with 4 half inch pipes.
After the model had been fabricated, the team moved forward to test out if the flow would evenly distributed in the aptly-named flow distribution system. To mimic the real tipping bucket, the team manually pour a 500mL water into the flow distribution system by hand, and looked at the distribution of the flow. Four small plastic bottles (each labeled as Quadrant 1, 2, 3 and 4) were put underneath each pipe to collect the amount of water that had flowed through each tube. The volume was then measured by pouring the contents of each bottle into a graduated cylinder.
The tipping bucket then was elevated so the small plastic bottles can be place beneath it. The elevation is done by placing the flow distribution system between two even holder as shown below.
Fig. 13 - Flow Distribution Set-up
Because the experiment is done by manually pouring water into the system, the team understood that it would be an error during this experiment. To anticipate this situation, the team decided to pour the water from each of the 4 sides of the system and see if what side it was poured from affected the volume of water collected in each bottle. Each side was done twice for a total of 8 trials.
** Note for the data analysis below: The volumes in the four bottles do not add up to the 500mL volume of the initial water because a small amount stayed in the top section. This happened because the shape of the plastic container has a slanted area that unintentionally collected the water.
As mentioned earlier the tipping bucket experiment, to test whether the water would be evenly distributed through the four pipes of the inlet system, UASB-R created a model with four partitions and tipped water from four different sides. A detailed top view of the model is shown in Figure 14.
Fig. 14 - A top view of the Model Water Distribution Tank.
The team, then collected data by measuring how much water went through each partition which is modeled by the code below.
print("hello")from aguaclara.play import *
from aguaclara.core.units import unit_registry as u
from aguaclara.core import physchem as pc
from aguaclara.core import constants as con
import numpy as np
import matplotlib.pyplot as plt
from aide_design.play import *
import math
#The list of Data Collected is shown below
#nomenClature: sidebuckettippedfrom+quadrantnumber = [1st trial, 2nd trial]
Front1st= [135, 107]*u.mL
Front2nd= [150, 119]*u.mL
Front3rd= [125, 134]*u.mL
Front4th= [70, 100]*u.mL
Back1st= [85, 88]*u.mL
Back2nd= [105, 102]*u.mL
Back3rd= [175, 147]*u.mL
Back4th= [125, 116]*u.mL
Left1st= [130, 107]*u.mL
Left2nd= [105, 91]*u.mL
Left3rd= [125, 133]*u.mL
Left4th= [70, 114]*u.mL
Right1st= [90, 112]*u.mL
Right2nd= [126, 110]*u.mL
Right3rd= [153, 152]*u.mL
Right4th= [93, 87]*u.mL
#The average of the two trials are taken to normalize the trial values so that they can be graphed and analyzed
#nomenClature: the initial of the side the water wastipped from+quadrant number
F1st = (Front1st[0]+Front1st[1])/2
F2nd = (Front2nd[0]+Front2nd[1])/2
F3rd = (Front3rd[0]+Front3rd[1])/2
F4th = (Front4th[0]+Front4th[1])/2
B1st = (Back1st[0]+Back1st[1])/2
B2nd = (Back2nd[0]+Back2nd[1])/2
B3rd = (Back3rd[0]+Back3rd[1])/2
B4th = (Back4th[0]+Back4th[1])/2
L1st = (Left1st[0]+Left1st[1])/2
L2nd = (Left2nd[0]+Left2nd[1])/2
L3rd = (Left3rd[0]+Left3rd[1])/2
L4th = (Left4th[0]+Left4th[1])/2
R1st = (Right1st[0]+Right1st[1])/2
R2nd = (Right2nd[0]+Right2nd[1])/2
R3rd = (Right3rd[0]+Right3rd[1])/2
R4th = (Right4th[0]+Right4th[1])/2
Horizontal_Array = ["1st", "2nd", "3rd", "4th"] #array used to indicate quadrant number
Front_Array = [F1st, F2nd, F3rd, F4th] #array containing the values of discharge
Back_Array= [B1st, B2nd, B3rd, B4th] #array containing the values of discharge
Left_Array= [L1st, L2nd, L3rd, L4th] #array containing the values of discharge
Right_Array= [R1st, R2nd, R3rd, R4th] #array containing the values of discharge
#Ploting the graph of water distribution from
index = np.arange(len(Horizontal_Array)) #this is used to assign the quadrant numbers to a variable which can be used to mark the horizontal axis of the bar graph
plt.bar(index,Front_Array) #graphs the plot
plt.xlabel('Quadrant Number', fontsize = 18)
plt.xticks(index, Horizontal_Array, fontsize=12, rotation=30)
plt.ylabel('Average Discharge (mL)', fontsize = 18)
plt.yticks(fontsize=12)
plt.title('Water Distribution Through Quadrants From The Front', fontsize = 20)
plt.show()
plt.savefig('./')
plt.bar(index,Back_Array)
plt.xlabel('Quadrant Number', fontsize = 18)
plt.xticks(index, Horizontal_Array, fontsize=12, rotation=30)
plt.ylabel('Average Discharge (mL)', fontsize = 18)
plt.yticks(fontsize=12)
plt.title('Water Distribution Through Quadrants From The Back', fontsize = 20)
plt.show()
plt.savefig('./')
plt.bar(index,Left_Array)
plt.xlabel('Quadrant Number', fontsize = 18)
plt.xticks(index, Horizontal_Array, fontsize=12, rotation=30)
plt.ylabel('Average Discharge (mL)', fontsize = 18)
plt.yticks(fontsize=12)
plt.title('Water Distribution Through Quadrants From The Left', fontsize = 20)
plt.show()
plt.savefig('./')
plt.bar(index,Right_Array)
plt.xlabel('Quadrant Number', fontsize = 18)
plt.xticks(index, Horizontal_Array, fontsize=12, rotation=30)
plt.ylabel('Average Discharge (mL)', fontsize = 18)
plt.yticks(fontsize=12)
plt.title('Water Distribution Through Quadrants From The Right', fontsize = 20)
plt.show()
plt.savefig('./')The code above normalizes the data attained from each quadrant and produces the graphs shown below.
Fig. 15 - The graph above represents the water distribution through the different quadrants when the water was tipped from the front.
Fig. 16 - The graph above represents the water distribution through the different quadrants when the water was tipped from the back.
Fig. 17 - The graph above represents the water distribution through the different quadrants when the water was tipped from the left.
Fig. 18 - The graph above represents the water distribution through the different quadrants when the water was tipped from the right.
According to the graphs above, it can be seen that for the tipping bucket tests done from the front and the left the distribution of water through the quadrants is fairly equal, but the distribution of water through the quadrants from the back and the right are skewed higher towards the third quadrant. To check whether this difference is significant a the data was modeled under a normalized curve, as can be seen below.
from aguaclara.play import *
from aguaclara.core.units import unit_registry as u
import numpy as np
from aide_design.play import *
import math
import pandas as pd
import seaborn as sns
%matplotlib inline
import statistics
#The list of Data Collected is shown below
#nomenClature: sidebuckettippedfrom+quadrantnumber = [1st trial, 2nd trial]
Front1st= [135, 107]*u.mL
Front2nd= [150, 119]*u.mL
Front3rd= [125, 134]*u.mL
Front4th= [70, 100]*u.mL
Back1st= [85, 88]*u.mL
Back2nd= [105, 102]*u.mL
Back3rd= [175, 147]*u.mL
Back4th= [125, 116]*u.mL
Left1st= [130, 107]*u.mL
Left2nd= [105, 91]*u.mL
Left3rd= [125, 133]*u.mL
Left4th= [70, 114]*u.mL
Right1st= [90, 112]*u.mL
Right2nd= [126, 110]*u.mL
Right3rd= [153, 152]*u.mL
Right4th= [93, 87]*u.mL
Volume_Array= [135, 107, 150, 119, 125, 134, 70, 100, 85, 88, 105, 102, 175, 147, 125, 116, 130, 107, 105, 91, 125, 133, 70, 114, 90, 112, 126, 110, 153, 152, 93, 87]
print (statistics.mean(Volume_Array))
print (statistics.stdev(Volume_Array))
VA = pd.DataFrame(Volume_Array) #makes the data readable by the sns package
sns.distplot(VA) #plots the data under a normal distribution
Modeling the data, Figure 19 shows that all the data is within 2 standard deviations(25 mL) of the mean (115 mL). Thus, it is understood that the data covers 95% of the normal bell curve, which shows a significant distribution in the data.
Fig. 19 - The normalized graph of the tipping bucket test data.
Qualitative and Quantitative:
Only two trials for each of the quadrants were done for each side of the inlet system, which is not ideal, and next semester the team will complete more trials; unfortunately time was extremely limited and UASB-R wasn't able to extensivly test for equal distribution.
However, in these small trials, UASB-R noticed two very important patterns that kept reoccuring.
- The two bottles further away from the side the water was being poured from obtained more liquid than the two bottles closer to that side.
For example: when the water was being poured from the back, the 3rd and 4th quadrants had more liquid and than 1st and 2nd. - In the eight trials the volumes were very different from each other.
From the modeling above, it is also understood that the differences in the data obtained are significant and that the water is not distributing fairly equally among the quadrants which doesn't ensure uniform distribution of the wastewater amongst the sludge blanket. With observation and data analysis pointing towards unequal distribution of water, the team concludes the following: either the tipping bucket must be angled with significant precision or the flow distribution system must be redesigned to account for two of the quadrants obtaining less water. Further, more it could also mean the inlet should split up into four opening later in the UASB reactor (i.e. in the tank).
This summer, a mix of UASB-D and UASB-R members from this semester will be staying over the summer to work on UASB. The summer UASB team will be building small-scale but fully functional UASB reactors at the Ithaca Area Wastewater Treatment Facility, using research done this semester and semesters past. What occurs over the summer will strongly direct what UASB-R chooses to work on in the fall of 2019. Some ideas have already been brought up, such as working on the dynamics of the sludge blanket: how long it takes to grow, how long does it take to settle after a pulse of water, etc., which could either involve experiments or researching these questions by going through literature. Continuing working on the flow distribution system is also on the table, but there is a currently a debate whether more testing would be worth it, or if the proposed design for the flow distribution system will be sufficient for an UASB reactor. But UASB-R will most likely focus on researching and solving problems that are discovered over the summer in the process of making and testing these UASB reactors. Therefore, a concrete, specific list of goals for UASB-R next semester will not be formed until later this summer.
The team has listed some key contacts for the UASB project so all members will know who to reach out to. They are summarized in the table below.
| Name | Role | |
|---|---|---|
| Dr. Ruth Richardson | Faculty Advisor | [email protected] |
| Ed Gottlieb | Wastewater plant operator, primary contact at Wastewater facility | [email protected] |
| N/A | Contact at Anheuser-Bush, source for anaerobic granules | N/A |
This article was developed under Assistance Agreement No. SU-83926801 awarded by the U.S. Environmental Protection Agency to Cornell University. It has not been formally reviewed by EPA. The views expressed in this document are solely those of [name of recipient or names of authors] and do not necessarily reflect those of the Agency. EPA does not endorse any products or commercial services mentioned in this publication.