Summer_2019_Final_Report.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Summer 2019 Final Report.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/AguaClara/Fluoride-Auto/blob/master/Summer_2019_Final_Report.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "awle31N5ykbr",
        "colab_type": "text"
      },
      "source": [
        "# Fluoride Removal, Summer 2019\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "#### Cindy Jin, Melissa Louie, Mike Reese\n",
        "#### 3 July 2019\n",
        "This publication Fluoride, Spring 2019 was developed under Assistance Agreement No. SU-83695001 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 Cindy Jin, Melissa Louie, and Mike Reese and do not necessarily reflect those of the Agency. EPA does not endorse any products or commercial services mentioned in this publication.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "eysmZuTJKOvY"
      },
      "source": [
        "## Table of Contents\n",
        "###[Abstract](#Abstract)\n",
        "<a href=\"#Abstract\"></a>\n",
        "\n",
        "###[Introduction](#Introduction)\n",
        "<a href=\"#Introduction\"></a>\n",
        "\n",
        "###[Literature Review](#Literature_Review)\n",
        "<a href=\"#Literature_Review\"></a>\n",
        "\n",
        "[Fluoride](#Fluoride)\n",
        "<a href=\"Fluoride\"></a>\n",
        "\n",
        "[Polyaluminum Chloride(PACl) and Defluoridation](#PACl)\n",
        "<a href=\"PACl\"></a>\n",
        "\n",
        "[Floc Blanket](#Floc_Blanket)\n",
        "<a href=\"Floc_Blanket\"></a>\n",
        "\n",
        "[Clay](#Clay)\n",
        "<a href=\"Clay\"></a>\n",
        "\n",
        "[Langmuir Isotherm](#Langmuir_Isotherm)\n",
        "<a href=\"Langmuir_Isotherm\"></a>\n",
        "\n",
        "###[Previous Work](#Previous_Work)\n",
        "<a href=\"Previous_Work\"></a>\n",
        "\n",
        "[Automated System](#Automated_System)\n",
        "<a href=\"Automated_System\"></a>\n",
        "\n",
        "[Gravity System](#Gravity_System)\n",
        "<a href=\"Gravity_System\"></a>\n",
        "\n",
        "###[Methods](#Methods)\n",
        "<a href=\"Methods\"></a>\n",
        "\n",
        "[Materials](#Materials)\n",
        "<a href=\"Materials\"></a>\n",
        "\n",
        "[Apparatus Changes](#Apparatus_Changes)\n",
        "<a href=\"Apparatus_Changes\"></a>\n",
        "\n",
        "[Sedimentation Tube](#Sedimentation_Tube)\n",
        "<a href=\"Sedimentation_Tube\"></a>\n",
        "\n",
        "[Gravity Powered System](#Grav_System)\n",
        "<a href=\"Grav_System\"></a>\n",
        "\n",
        "[Automated System](#Automated_System)\n",
        "<a href=\"Auto_System\"></a>\n",
        "\n",
        "###[Procedure](#Procedure)\n",
        "<a href=\"Procedure\"></a>\n",
        "\n",
        "###[Results and Analysis](#Results_and_Analysis)\n",
        "<a href=\"Results_and_Analysis\"></a>\n",
        "\n",
        "[Consistent Flocculation in the Gravity System](#Consistent_Flocculation)\n",
        "<a href=\"Consistent_Flocculation\"></a>\n",
        "\n",
        "[Determining an Optimal Upflow Velocity](#Upflow_Velocity)\n",
        "<a href=\"Upflow_Velocity\"></a>\n",
        "\n",
        "[Bottom Geometry](#botgeo)\n",
        "<a href=\"botgeo\"></a>\n",
        "\n",
        "[Troubleshooting Gel](#Troubleshooting_Gel)\n",
        "<a href=\"Troubleshooting_Gel\"></a>\n",
        "\n",
        "[Sedimentation Tube Calculations](#Calculations)\n",
        "<a href=\"Calculations\"></a>\n",
        "\n",
        "[Sedimentation Tube Design](#Sed_Tube_Design)\n",
        "<a href=\"Sed_Tube_Design\"></a>\n",
        "\n",
        "[Introducing Clay into the System](#Introducing_Clay)\n",
        "<a href=\"Introducing_Clay\"></a>\n",
        "\n",
        "[Increasing the Size of the Floc Weir](#Floc_Weir)\n",
        "<a href=\"Floc_Weir\"></a>\n",
        "\n",
        "[Testing the Limits of the Fluoride Probe](#Fluoride_Probe)\n",
        "<a href=\"Fluoride_Probe\"></a>\n",
        "\n",
        "[Fluoride Concentration and Turbidity over Time](#Fluoride_Turbidity)\n",
        "<a href=\"Fluoride_Turbidity\"></a>\n",
        "\n",
        "[Floc Blanket Height and Fluoride Concentration](#Blanket_Concentration)\n",
        "<a href=\"Blanket_Concentration\"></a>\n",
        "\n",
        "[Testing a Horizontal Reactor](#Horizontal_Reactor)\n",
        "<a href=\"Horizontal_Reactor\"></a>\n",
        "\n",
        "[Fitting the Experimental Data to the Isotherm](#Isotherm)\n",
        "<a href=\"Isotherm\"></a>\n",
        "\n",
        "[Molecular Interactions Between PACl and Fluoride](#Molecular)\n",
        "<a href=\"Molecular\"></a>\n",
        "\n",
        "[Jar Test](#jartest)\n",
        "<a href=\"jartest\"></a>\n",
        "\n",
        "[Implementing the Jar Test in a Continuous System](#Continuous_System)\n",
        "<a href=\"Continuous_System\"></a>\n",
        "\n",
        "###[Conclusions](#Conclusions)\n",
        "<a href=\"Conclusions\"></a>\n",
        "\n",
        "###[Future Work](#Future_Work)\n",
        "<a href=\"Future_Work\"></a>\n",
        "\n",
        "###[Bibliography](#Bibliography)\n",
        "<a href=\"Bibliography\"></a>\n",
        "\n",
        "###[Manual](#Manual)\n",
        "<a href=\"Manual\"></a>\n",
        "\n",
        "###[Python Code](#Python_Code)\n",
        "<a href=\"Python_Code\"></a>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lGzNCdLa1f38",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Abstract\"></a>\n",
        "##Abstract\n",
        "The Summer 2019 Fluoride team investigated issues related to the Automated and Gravity systems, with the goal of determining ideal conditions for removing fluoride. Initial qualitative tests prompted fabrication changes, such as an increase in the plate settler angle in the sedimentation tube. Jar testing was also done to examine if a floc blanket is needed for fluoride removal. Several changes in the sedimentation tube were tested to develop an optimal system for fluoride removal. Fluoride concentrations were measured with an ion electrode to determine removal efficiency and to continue developing the Langmuir Adsorption Isotherm model. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TpShT1NO1jsq",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Introduction\"></a>\n",
        "## Introduction\n",
        "\n",
        "The majority of AguaClara plants obtain their source of water from surface water, originating from rivers and mountain streams. AguaClara's mission of providing access to safe drinking water for all has motivated the AguaClara team to expand and develop new goals to provide safe drinking water from groundwater sources. By considering and analyzing the differences in drinking water collected from groundwater sources in other regions around the world and the needs associated with these sources, AguaClara can serve a broader range of communities around the world who are deprived of access to safe drinking water.\n",
        "\n",
        "At some locations where there is little to no fluoride in drinking water, fluoride is added in for the purpose of strengthening teeth and deteriorating natural deposits [(NYC DEP, 2016)](http://www.nyc.gov/html/dep/pdf/wsstate16.pdf). However, there are detrimental consequences when there is an excess concentration of fluoride in the water being consumed by inhabitants in the area. For example, there has been moderate to high fluoride pollution in regions such as Algeria, Thailand, Ghana, India, and Iran, whose major source of drinking water comes from groundwater sources. The pollution of groundwater in these regions is due to the mobile dissolution of fluorite, apatite, and topaz from weathered rocks, increasing the levels of fluoride in the groundwater [(Bhattacharya and Samal, 2018)](http://www.isca.in/rjrs/archive/v7/i4/6.ISCA-RJRS-2018-028.pdf). The effects of consuming excess fluoride include dental, skeletal, developmental, neurological, endocrine, reproductive, and carcinogenic issues that affect the health of the communities in areas of high fluoride concentrations [(Bhattacharya and Samal, 2018)](http://www.isca.in/rjrs/archive/v7/i4/6.ISCA-RJRS-2018-028.pdf). Therefore, the World Health Organization (WHO) regulated the concentration limit for fluoride in drinking water to be 1.5 mg/L [(Water-related diseases, 2016)](https://www.who.int/water_sanitation_health/diseases-risks/diseases/fluorosis/en/). However, more than 65 million Indians consume drinking water above the recommended limit of 1.5 mg/L and some locations in India contain up to 20 mg/L, which risks the health of many communities [(LeChevallier and Au, 2004)](https://www.who.int/water_sanitation_health/publications/9241562552/en/).\n",
        "\n",
        "The AguaClara team aspired to achieve the recommended limit of 1.5 mg/L or lower stated by the World Health Organization. The gravity system was the one tested with the implications that it could one day be implemented on-site for fluoride removal in the communities with excess concentrations of fluoride. The automated system was tested moreso to investigate optimal conditions for fluoride removal in an AguaClara reactor and the feasibility of the design. Nevertheless, the Fluoride team tested both the automated apparatus and the gravity-powered apparatus with PACl and fluoride to optimize fluoride removal to achieve residual concentration of less than 1.5 mg/L. If the gravity-powered fluoride removal apparatus were to operate successfuly in the optimal conditions of removing fluoride to a concentration of 1.5 mg/L, feasibility analysis on site would be performed to test whether the system can be used in the communities in India, which would allow AguaClara to provide access to safe drinking water in India and other locations that obtain water from groundwater sources."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iPzGkI2f4-wr",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Literature_Review\"></a>\n",
        "## Literature Review\n",
        "\n",
        "<a name=\"Fluoride\"></a>\n",
        "\n",
        "###Fluoride\n",
        "Groundwater in some developing countries, including India, is contaminated by industrial waste and geological deposits that often contain fluoride. If ingested in high concentrations over an extended period of time, fluoride can pose serious health concerns. While fluoride concentrations of 1.0 mg/L to 1.5 mg/L in drinking water are generally considered safe and can even help improve dental enamel in children, levels above 1.5 mg/L can cause dental and skeletal fluorosis ([Jagtap et al, 2012](https://pubs.acs.org/doi/pdf/10.1021/cr2002855)). \n",
        "\n",
        "Fluoride levels are generally found to be below 1.5 mg/L in seawater, rivers, and lakes, but tend to be much higher in groundwater sources, especially in areas where minerals containing fluoride are located in abundance ([Fawell et al, 2006](https://books.google.com/books?id=7xu1yf-OC8oC&lpg=PR5&ots=f1CYqkjcYO&dq=fluoride%20drinking%20water&lr&pg=PA6#v=onepage&q=fluoride%20drinking%20water&f=false)). Thus, the presence of fluoride poses an especially significant risk in regions that obtain much of their drinking water from underground aquifers. In rural India, for example, approximately 80% of water used for drinking comes from groundwater sources ([Jagtap et al, 2012](https://pubs.acs.org/doi/pdf/10.1021/cr2002855)). A study by Brindha et al. that was carried out in India’s Nalgonda district found groundwater concentrations of fluoride as high as 8.8 mg/L, and determined that only around 52% of the water sampled was safe to drink ([2009](https://link.springer.com/article/10.1007/s10661-010-1348-0)).\n",
        "\n",
        "The World Health Organization has set a limit on fluoride in drinking water at 1.5 mg/L based on evidence that adverse health effects become prevalent at concentrations above this standard. Furthermore, the WHO notes that in regions with elevated levels of fluoride intake via drinking water, \"fluoride is a significant cause of morbidity.\" ([WHO, 2004](https://apps.who.int/iris/handle/10665/42852)). Therefore, the AguaClara Fluoride team aims to reduce effluent fluoride to a concentration below this standard of 1.5 mg/L.\n",
        "\n",
        "<a name=\"PACl\"></a>\n",
        "###Polyaluminum Chloride (PACl) and Defluoridation\n",
        "One method of defluoridation for drinking water is a system that implements coagulation, flocculation, and sedimentation. Coagulation is a process that involves adding a positively charged chemical coagulant to the raw, untreated water. Once the coagulant has been sufficiently flocculated, or mixed, it begins to destabilize colloidal and dissolved particles in the raw water ([EPA, 2007](https://oaspub.epa.gov/tdb/pages/treatment/treatmentOverview.do?treatmentProcessId=1934681921)). The coagulant that is often used for defluoridation of water is polyaluminum chloride, or PACl, with the chemical formula Al2(OH)3Cl3 ([Gebbie, 2001](http://wioa.org.au/conference_papers/2001/pdf/paper6.pdf)). PACl is a better choice of coagulant in comparison to other common coagulants such as aluminum sulfate, or alum, because of its ability to treat water in a larger pH range of around 5.0 to 8.0 in contrast to alum’s effective pH range of around 5.5 to 6.5. Additionally, the absence of sulfate in PACl prevents the buildup of hydrogen sulfide in wastewater, which has been known to corrode pipes and other components of the sewer system. Both PACl and alum lead to residual aluminium in the effluent water, which must then be removed to meet the 0.2 mg/L standard set by the Environmental Protection Agency. The usage of PACl, however, tends to result in less residual aluminium, which further supports using PACl as the coagulant for fluoride removal instead of alum ([Dubey et. al, 2018](https://www.sciencedirect.com/science/article/pii/S0167732218303490)).\n",
        "\n",
        "\n",
        "After PACl is added to raw water, the molecules attract dissolved fluoride ions and the mixture enters the flocculation stage of treatment. In flocculation, the destabilized particles from coagulation are encouraged to agglomerate together to form \"flocs\". There are several ways to instigate the agglomeration of these particles, with some common methods being a gentle mix, or a baffled plug-flow reactor ([EPA, 2007](https://oaspub.epa.gov/tdb/pages/treatment/treatmentOverview.do?treatmentProcessId=1934681921)). The goal during the period of flocculation is to maximize the probability of particle interaction between destabilized matter so that these species of interest will form flocs in the reactor that have a higher density than the water around them. To reiterate, \"Aggregation requires successful collisions. Success is defined by particles attaching when they collide\" ([Weber-Shirk et al, 2018](aguaclara.github.io/Textbook/index.html)). These successful collisions allow for the removal of fluoride in the sedimentation process, where the agglomerated particles are separated from the water using gravity settling or other clarification methods ([EPA 2007](https://oaspub.epa.gov/tdb/pages/treatment/treatmentOverview.do?treatmentProcessId=1934681921)). \n",
        "\n",
        "<a name=\"Floc_Blanket\"></a>\n",
        "\n",
        "###The Floc Blanket\n",
        "\n",
        "Floc blankets are formed in tanks where water is fed downwards into a hopper-shaped geometry displayed in Figure 1. The water is then redirected upwards, which resuspends floc particles that have settled to the bottom of the tank so they can agglomerate into a fluidized layer of flocs in the tank ([Garland et al, 2016](http://doi.org/10.1089/ees.2015.0314). This collection of flocs, or blanket, typically rises to a point that is controlled by some type of weir or other means of particle separation from the treated water ([Gregory 1977](https://spiral.imperial.ac.uk/bitstream/10044/1/22645/2/Gregory-R-1977-MPhil-Thesis.pdf)).\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://aguaclara.github.io/Textbook/_images/diffuser_jetreverser.png\"height=\"200\">\n",
        "  \n",
        "  **Figure 1:** The inlet jet geometry that redirects water upwards and allows the formation of a floc blanket.\n",
        "  </center>\n",
        "  \n",
        "  <center>\n",
        "  <img src=\"https://aguaclara.github.io/Textbook/_images/Floc_Blanket_Floc_Hopper.png\" height = \"300\">\n",
        "  \n",
        "  **Figure 2:** A side view of a sedimentation tank. Once the floc blanket builds to the height of the floc weir on the left side of the image, excess flocs overflow into the weir instead of continuing to increase the height of the floc blanket.\n",
        "  </center>\n",
        "\n",
        "Using a floc blanket for particle removal is beneficial because, as water travels upwards through the blanket, there is an increased chance of collisions between remaining particles in the water and the particles in the floc blanket; particles in the water are then more likely to be captured by the floc blanket instead of escaping with the clean effluent. While this process greatly increases removal efficiency, it should be noted that the floc blanket does not capture all particles, because of shear forces acting on smaller flocs that prevent them from remaining attached to larger flocs.\n",
        "\n",
        "The stability and quality of the floc blanket depends on both the settling velocity of flocs and the upflow velocity through the sedimentation tank ([Su et al, 2004](http://www.sciencedirect.com/science/article/pii/S0043135403005025)). For this reason, it is important that these two aspects are controlled carefully. Studies have shown that when one of these parameters experiences fluctuations, floc blankets can fail to function as intended, which may subsequently lead to poor effluent water quality ([Su et al, 2004](http://www.sciencedirect.com/science/article/pii/S0043135403005025)). In one case, a clarifier in Taiwan lost its entire floc blanket during a period of low turbidity measurements in the influent raw water ([Lin et al, 2004](https://ascelibrary.org/doi/10.1061/%28ASCE%290733-9372%282004%29130%3A12%281481%29)). Furthermore, the conditions must be maintained to allow for the resuspension of agglomerated flocs into the fluidized bed, but not to a superfluous degree where flocs would be torn apart by high shear forces ([Garland et al, 2016](http://doi.org/10.1089/ees.2015.0314)).\n",
        "\n",
        "<a name=\"Clay\"></a>\n",
        "\n",
        "###Clay\n",
        "\n",
        "Kaolinite clay was introduced into the fluoride system to investigate its usefulness as an indicator for flocculation. When fluoride flocs form with PACl, the resulting solids are clear and can only be seen when closely examining the sedimentation tube with a flashlight. It was hypothesized that by introducing clay after the flocculator, there would be some interaction between the fluoride flocs and the clay that would involve fluoride flocs adhering to clay particles and thereby becoming more visible. Furthermore, flocs that contain clay and PACl tend to be larger than those containing fluoride and PACl, which was thought to improve removal of the flocs during sedimentation ([Savenko 1999](https://link.springer.com/content/pdf/10.1023%2FA%3A1010496606751.pdf)). Research has shown that clay can also play a role - especially in groundwater applications - in the adsorption of fluoride and other pollutants of interest ([Savenko 1999](https://link.springer.com/content/pdf/10.1023%2FA%3A1010496606751.pdf)). Savenko claims that, when paired with certain aluminum coagulation processes, each gram of clay added has been able to absorb up to 3.3 mg of fluoride ([1999](https://link.springer.com/content/pdf/10.1023%2FA%3A1010496606751.pdf)).\n",
        "\n",
        "<a name=\"Langmuir_Isotherm\"></a>\n",
        "\n",
        "###Langmuir Isotherm\n",
        "\n",
        "The Langmuir isotherm is a model used to describe the adsorption process, in which a chemical species of interest (the adsorbate) is attracted to the surface of a solid (adsorbent). The two then attach by a physical or chemical bond ([Foo and Hameed, 2010](https://doi.org/10.1016/j.cej.2009.09.013)). These models allow for the delineation of relationships between the pollutant of interest and the adsorbent being used, and are thus very useful in giving a visual representation of adsorption effiency based on specific experimental conditions ([Foo and Hameed, 2010](https://doi.org/10.1016/j.cej.2009.09.013)).\n",
        "\n",
        "The Langmuir isotherm is governed by the following equation [(Washington Faculty)](http://mimoza.marmara.edu.tr/~zehra.can/ENVE401/3.%20Adsorption%20Equilibria.pdf).\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Isotherm%20Equation.png\" height = \"75\">\n",
        "  </center>\n",
        "  \n",
        "  *qA = Uptake (mg Fluoride/g PACl)*\n",
        "\n",
        "*QM = Maximum adsorption density (mg/L Fluoride)*\n",
        "\n",
        "*Kad = Langmuir constant*\n",
        "\n",
        "*CA = Effluent Fluoride Concentration (mg/L Fluoride)*\n",
        "\n",
        "\n",
        "Past AguaClara fluoride teams used an isotherm to model the relationship between the mass of fluoride absorbed (mg) per gram of PACl used and the effluent concentration of fluoride (mg/L). In this instance, fluoride is the pollutant of interest, and PACl can be considered the adsorbent material. The most recent version of this isotherm model can be seen below in Figure 3.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Langmuir%20Isotherm%20Data/isotherm.png\">\n",
        "  \n",
        "**Figure 3:**  Langmuir isotherm model developed by the Spring 2019 Fluoride Automated team.\n",
        "</center>\n",
        "\n",
        "The goal of using the isotherm is to optimize the uptake of fluoride, or efficiency of the PACl. Ideally, this model would indicate an accurate dosage of PACl for a detected influent concentration of fluoride that could lower the effluent fluoride concentration to a level below the WHO standard of 1.5 mg/L for drinking water.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LuA45vR6cCGW",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Previous_Work\"></a>\n",
        "## Previous Work"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vn8QjMw95Wga",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Automated_System\"></a>\n",
        "\n",
        "### Automated System\n",
        "The AguaClara Fluoride Automated System began in Fall 2016 after receiving an EPA Phase II grant in Spring 2016 [(Cheng et. al., 2016)](https://drive.google.com/file/d/0Bzej1vulT_XyWHZyLXdIWlk1RkE/view). The team focused on using PACl for particle removal using only a floc blanket. The team ran tests with red dye in place of fluoride so that turbidity and visual observation could be used to gauge particle removal. The team determined that the floc blanket must be at some minimum height greater than 5 cm for effective removal of red dye. In addition, a sloped bottom geometry for the base of the sedimentation tube was fabricated to prevent the settling of flocs at the bottom of the reactor and subsequent clogging. The main goal of this new bottom geometry was to remove any flat horizontal surfaces so that flocs that reached the bottom would have to recirculate instead of settling. \n",
        "\n",
        "The 2017 teams used clay in the system, but stopped using clay after observing high levels of turbidity in the effluent. Using high levels of PACl (25 mg/L), the team was able to produce an effluent with a concentration of as low as 0.05 mg/L [(Longo et. al., 2017)](https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202017/Fluoride%20Spring%202017.pdf). Additionally, initial tests were conducted with both single and multi-reactor systems, but the single reactor system was ultimately chosen after taking large-scale cost and labor into consideration [(Longo et. al., 2017)](https://github.com/AguaClara/Fluoride-Auto/blob/master/Fall%202017/fluoride-fall-2017.pdf).\n",
        "\n",
        "The Spring 2018 team ran experiments in which they either changed the influent PACl concentration or the influent fluoride concentration; in order to collect data for the developing Langmuir Isotherm, an adsorption model was thought of to describe the interaction between polyaluminum chloride and particles in the water such as red dye or fluoride ions [(Akpan et. al., 2018)](https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202018/FluorideReportSp18.md). Skepticism over the accuracy of the probe led to somewhat inconclusive data.\n",
        "\n",
        "In Summer 2018, the team concentrated on continuing to run experiments to gain more data to plot against the Langmuir Isotherm [(Pang et. al., 2018)](https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202018/Fluoride_Summer2018.md). The team plotted uptake, or the mass of fluoride removed per the mass of influent PACl, against the effluent fluoride concentration and concluded that PACl was effective in removing fluoride from the influent water, even if an optimal dosage of PACl had not yet been determined.\n",
        "\n",
        "The Fall 2018 team experienced the significant obstacle of gel buildup in the reactor [(Mehrabyan et. al., 2018)](https://github.com/AguaClara/Fluoride-Auto/blob/master/Fall%202018/Automated%20System/Fall_2018_Report.md). Functioning floc blankets should be a fluidized bed of particles that are constantly recirculating, while gel is a mass of flocs that tend to adhere to the side of the reactor and do not recirculate. The reoccurrence of gel in most of the experiments caused early failure due to blockage of the floc weir entrance and the ensuing escape of flocs through the effluent tubing.\n",
        "\n",
        "The Spring 2019 team continued the task of testing different ratios of PACl to fluoride and plotting the data of previous semesters against the Langmuir Isotherm [(Sausele et. al., 2019)](https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202019/Fluoride-Auto-Spring-2019-Report.md). The team added a flow accumulator in the system to prevent pulsing. The data from Spring 2018 to Spring 2019 is shown in the Results and Analysis section. The Spring 2019 team also transitioned from using MathCAD for determining experimental flow rates to using Python, which is currently being used for experimental calculations as well as graphing and data analysis. \n",
        "\n",
        "Additionally, the Spring 2019 team ran tests to see if the fluoride probe could detect fluoride in the form of flocs. High-turbidity samples of effluent that contained flocs were measured with the fluoride probe. After filtering the samples through mesh, their concentrations were measured again. The team found significant percent increases in measured fluoride concentration after filtering, which led to the conclusion that the mesh helped break up the flocs into a form that the probe was able to detect."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m2BcoJ085Ywr",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Gravity_System\"></a>\n",
        "\n",
        "### Gravity System\n",
        "\n",
        "\n",
        "The gravity powered apparatus started developing in the Fall 2017 and Spring 2018 team using the concept of constant head tanks [(Akpan et al., 2017)](https://github.com/AguaClara/Fluoride-Auto/blob/master/FluorideReportSp18.md). These constant head tanks control flow rates through the system: the Fluoride Constant Head Tank drives the upflow velocity in the sedimentation tank while the PACl Constant Head Tank drives the PACl flow rate and dosage. The PACl and fluoride constant head tanks have float valves that allow for a constant volumetric flow through the system, which can then be calculated by measuring the output volume over time [(Pang et al., 2018)](https://github.com/AguaClara/Fluoride_Gravity/blob/master/Fall%202018/Fluoride_Grav_Fall2018_Report.md).\n",
        "\n",
        "The Summer 2018 team improved the system by making the gravitational energy easily adjustable with sliders on the PACl Stock Tank, PACl Constant Head Tank, and effluent pipe, which allowed for easier control and measurement for flow through the system (Figure 4). \n",
        "\n",
        "<center>\n",
        "  \n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride_Gravity/master/Spring%202019/Diagram.PNG\" height = 600 width = 400>\n",
        "  \n",
        "  **Figure 4:** The Summer 2018 gravity-powered bench with adjustable sliders for stock tanks.\n",
        "  </center>\n",
        "\n",
        "In Fall 2018, the Fluoride Gravity team was officially established. The priority of the Fall 2018 semester was to find a way to easily measure the flow rate of coagulant free-falling into the fluoride system without any headloss. The team used microtubing to add PACl into an IV drip chamber, but head loss still occurred [(Pang et al., 2018)](https://github.com/AguaClara/Fluoride_Gravity/blob/master/Fall%202018/Fluoride_Grav_Fall2018_Report.md).\n",
        "\n",
        "The Spring 2019 team created several different ways to introduce PACl into the system without any head loss and to effectively mix the PACl and fluoride. Since fluoride flocs are nearly invisible, the team tested the gravity powered system with red dye and PACl. The PACl was added directly into the Fluoride Stock Tank using first a T-junction, then a clear horizontal T-junction, and finally a vertical clear T-junction. The Spring 2019 team successfully determined that the vertical clear T-junction modification was the optimal method of introducing PACl into the system in a way that ensured consistent flocculation and effective mixing [(Huang et al., 2019)](https://github.com/AguaClara/Fluoride_Gravity/blob/master/Spring%202019/Spring%202019%20Report.md). The updated Spring 2019 schematic is shown below in [Figure 8](#Figure_8)<a href=\"#Figure_8\"></a>.\n",
        "\n",
        "The goal of the Summer 2019 Fluoride team for the gravity system was to run tests with fluoride in the gravity system and obtain data regarding efficiency of the system."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HNATXIDfa_RX",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Methods\"></a>\n",
        "##Methods\n",
        "\n",
        "<a name=\"Experimental Apparatus\"></a>\n",
        "## Experimental Apparatus"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ToB7a10c8Ubg",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Materials\"></a>\n",
        "\n",
        " ### Materials\n",
        " * One 600 RPM pump and three 100 RPM pumps\n",
        " * Transparent 2.54 cm (1\") PVC piping\n",
        " * Flexible and hard 0.635 cm (1/4\") tubing and Microbore tubing\n",
        " * One Turbidimeter\n",
        " * One Fluoride Ion Selective Electrode (ISE) Probe (VWR International, Catalog No. BK511141 (EA))\n",
        "  * Access the general manual [here](https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202019/Resources/Fluoride%20Probe%20Manual.pdf).\n",
        "  * Access the comprehensive manual [here](https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202019/Resources/Fluoride%20Selective%20Electrode%20Comprehensive%20Manual.pdf).\n",
        " * 1\" and 3/4\" clear Polyvinyl Chloride, or PVC (McMaster Carr)\n",
        " * One clear Size 1 tee (McMaster Carr)\n",
        " * One Bottom Geometry (High Rate Sedimentation Bottom Geometry team)\n",
        " * One 47 mm In-Line Filter Holder (Pall Corporation)\n",
        " * 1000 mg/L Polyaluminum Chloride (PACl) Stock Solution\n",
        " * 1000 mg/L Fluoride Stock Solution\n",
        " * 1000 mg/L Clay Stock Solution\n",
        " * Connectors and buckets for stock tanks\n",
        " * Two bottles for fluoride stock and flow accumulator\n",
        " * Three stir plates with stir bars for the stock solutions"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7T8RE0_rWixT",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Apparatus_Changes\"></a>\n",
        "### Apparatus Changes\n",
        " * **Addition of the Flow Accumulator in the Automated System:** Early on in the summer, the team observed pulsing in the influent water tube, which continued in the flocculator and sedimentation tube. Thus, the team implemented the Flow Accumulator, which is a bottle with both an influent and effluent opening. The bottle is filled with water just enough to cover these openings, then capped tightly in order to allow the air pressure inside the bottle to facilitate the flow of water out of the bottle, free of pulsations from the pump.\n",
        " * **Increased Angle of the Sedimentation Tube:** In previous semesters, the team used a sedimentation tube with a bend that was 60° from the horizontal. In the initial tests that the Summer 2019 team ran, flocs seemed to be accumulating above the floc weir, on the plate settlers. The team then increased the angle of the bend to 70° in an attempt to allow flocs to slide down the plate settlers more easily and thus prevent the build up of flocs above the floc weir.\n",
        " * **Uniform Reactor for Automated and Gravity System:** In the Spring 2019 semester, the Automated and Gravity systems used sedimentation tubes that were uniform in angle (the bend was 60° from the horizontal) and lengths. However, the systems differed in their placement of the floc weir. While the Automated system floc weir was on the vertical part of the sedimentation tube (Figure 5),  the Gravity system floc weir was above the bend of the tube (Figure 6). The Automated team also connected a 1 RPM waste pump to the floc weir to remove flocs. The Summer 2019 decided to use two sedimentation tubes with the same specifications in order to be able to compare experiments done on both systems. Thus, the team fabricated two sedimentation tubes with a bend that was 70° from the horizontal and a floc weir above the bend instead of on the vertical part of the sedimentation tube. The Summer 2019 team also decided not to include the waste pump in the system, as gravity proved to be sufficient for flocs to exit the weir. Finally, the Spring 2019 Automated and Gravity teams used different bottom geometries, but after initial testing, a working bottom geometry was determined and implemented in both systems. See [Choice of Bottom Geometry in Sedimentation Tube](#Choice_of_Bottom_Geometry_in_Sedimentation_Tube)<a href=\"Choice_of Bottom_Geometry_in_Sedimentation Tube\"></a> for more information.\n",
        " \n",
        "<center>\n",
        " <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Images/Sed%20Tube%20Measurements.JPG\"height=\"300\">\n",
        "  \n",
        "**Figure 5:**  The sedimentation tube from the Spring 2019 Automated team, with the floc weir on the vertical portion of the tube.\n",
        " \n",
        "  </center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ExfbOSIEeY8d",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Sedimentation_Tube\"></a>\n",
        "### Sedimentation Tube\n",
        "<center>\n",
        " <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Sedimentation%20Tube.png\" height=\"500\">\n",
        " \n",
        "**Figure 6:**  The Summer 2019 sedimentation tube.\n",
        "   </center>\n",
        " \n",
        " Figure 6 shows the specifications of the new sedimentation tube that was fabricated. The angle of the plate settler from the vertical section of the apparatus was increased from 150° to 160°. This change was made because initial tests revealed fluoride flocs creating a non-fluidized gel substance that would accumulate on the plate settler instead of sliding down and falling through the floc weir. By increasing the angle of the tube, the downward force of gravity acting on flocs above the floc weir increased."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V2HNrGxyWON9",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Grav_System\"></a>\n",
        "\n",
        "###Gravity Powered System\n",
        "<center>\n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride_Gravity/master/Spring%202019/Schematic.PNG\" height = 500 width = 600 >\n",
        "  \n",
        "**Figure 7:**  A diagram of the Spring 2018 gravity-powered system with height differentials driving the system.\n",
        "</center>\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vr81h0E6cGNS",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Figure_8\"></a>\n",
        "<center>\n",
        "\n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride_Gravity/master/Spring%202019/current%20drawing.png\" height = 550 width = 550> \n",
        " \n",
        "\n",
        "**Figure 8:** The updated Summer 2019 lab bench for the Fluoride gravity apparatus.\n",
        "</center>\n",
        "\n",
        "### Process Flow Through the Reactor\n",
        "1. Fluoride stock solution from the Fluoride Stock Tank flows to the Fluoride Constant Head Tank.\n",
        "2. The fluoride stock solution flows to the Vertical T-junction while PACl flows from the PACl Constant Head Tank to the Vertical T-junction.\n",
        "3. The PACl and fluoride stock solution mix in the Vertical T-junction.\n",
        "4. The PACl and fluoride mixture flows through the Flocculator.\n",
        "5. The flocculated mixture flows to the Sedimentation Tank.\n",
        "6. A floc blanket forms and grows to the height of the floc weir. \n",
        "7. Flocs from the floc blanket spill into the Floc Weir. At the same time, flocs that passed the height of the floc blanket fall onto the Plate Settler and slide down until they reach the height of the Floc Weir and fall into it.\n",
        "8. Clean effluent water exits through the top of the Sedimentation Tube and flows back downwards until it exits the plastic tubing."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SwWV53fwb2XD",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Auto_System\"></a>\n",
        "### Automated System\n",
        "#### General System\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/General%20Lab%20Bench.png\" height=\"300\">\n",
        "  \n",
        "  **Figure 9:** The Summer 2019 Automated lab bench setup.\n",
        "  </center>\n",
        "  \n",
        "  <center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/General%20Schematic.png\" height=\"300\">\n",
        "  \n",
        "  **Figure 10:** A schematic of the Summer 2019 Automated system showing the flow of water through the reactor.\n",
        "  </center>\n",
        "\n",
        "### Process Flow Through Reactor\n",
        " 1. Water flows from the tap to the Flow Accumulator, then continues on through the main tubing.\n",
        " 2. Fluoride solution flows from the F- Stock tank to the main tubing, where it becomes diluted by the stream of water.\n",
        " 3. PACl flows from the PACl Stock tank to main tubing, where it is added to the fluoride and tap water making a PACl-fluoride mixture.\n",
        " 4. The mixture travels through the Flocculator.\n",
        " 5. The mixture flows into the Sedimentation Tube.\n",
        " 6. The floc blanket grows to the height of the Floc Weir, and the overflow of flocs exits the Sedimentation Tube through the Floc Weir. Flocs that do not initially exit through the Floc Weir fall onto the Plate Settlers and slide down into the Floc Weir. Meanwhile, the effluent stream exits the top of the Sedimentation Tube.\n",
        " 7.  The Turbidimeter measures the turbidity of the effluent stream.\n",
        " 8. The mixture flows through the Mesh Filter. See [Testing the Limits of the Fluoride Probe](#Testing_the_Limits_of_the_Fluoride_Probe)<a href=\"#Testing_the_Limits_of_the_Fluoride_Probe\"></a> for more information.\n",
        " 9. The Fluoride Probe measures the effluent fluoride concentration.\n",
        " 10. The effluent stream flows into the sink.\n",
        " \n",
        "#### Clay System\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Lab%20Bench.png\" height=\"300\">\n",
        "  \n",
        "  **Figure 11:** The Automated lab bench with the addition of clay after the flocculator.\n",
        "       </center>\n",
        "  \n",
        "  <center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Schematic.png\" height=\"300\">\n",
        "  \n",
        "**Figure 12:**  A schematic of the Automated system with the addition of clay.\n",
        "    </center>\n",
        "    \n",
        "#### Increased Floc Weir Size\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Large%20Floc%20Weir.png\" height=\"300\">\n",
        " \n",
        "**Figure 13:**  The sedimentation tube with a 1\" inner diameter floc weir.\n",
        "  </center>\n",
        "    \n",
        "#### Lengthened Sedimentation Tube\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/long_sed.jpg\" height=\"300\">\n",
        "  \n",
        "**Figure 14:** Long sedimentation tube layout to test for increased height of floc blanket. The angle between the plate settler and floc weir has since been increased.\n",
        "    </center>\n",
        "  \n",
        "#### Horizontal Reactor\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Labeled%20Horizontal%20Reactor.png\" height=\"200\">\n",
        "  \n",
        "  **Figure 15:** The sedimentation tube that is oriented horizontally and develops no floc blanket.\n",
        "  </center>\n",
        "    \n",
        "#### Two-Reactor System\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Two%20Sed%20Tube%20Lab%20Bench.JPG\" height=\"300\">\n",
        "  \n",
        "  **Figure 16:** The system with two sedimentation tubes in series.\n",
        "  </center>\n",
        "  \n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Two%20Reactor%20Schematic.png\" height=\"300\">\n",
        "  \n",
        "  **Figure 17:** A schematic of the setup with two sedimentation tubes in series.\n",
        "  </center>\n",
        "    "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "14QXCacy8axM",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Procedure\"></a>\n",
        "## Procedure"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BHgjvib_M_Ho",
        "colab_type": "text"
      },
      "source": [
        "### Gravity System \n",
        "1. Make 4-8 L of a desired concentration of fluoride stock solution in the Fluoride Stock Tank. More stock solution may need to be added during the experiment if it runs out, depending on the time of the experiment and flow rate through the reactor.\n",
        "2. Adjust the height of the tray that holds the PACl Constant Head Tank. Volumetrically measure the flow rate (collect a sample from the microtubing in a graduated cylinder and use a timer to see how long the sample collection takes) and continue to adjust the tray based on the desired flow rate.\n",
        "3. Open the ball valve between the Fluoride Stock Tank and Fluoride Constant Head Tank and allow the Fluoride Constant Head Tank to its maximum capacity.\n",
        "4. Open the ball valve between the Fluoride Constant Head Tank and the Vertical T-junction. Allow the Vertical T-junction to fill; adjustment of the Vertical T-junction may be necessary so that the level of water inside it stays below the level of its influent valve.\n",
        "5. Place the end of the microtubing for the PACl inside the Vertical T-junction (above the water). Open the ball valve directly below the PACl Constant Head Tank and make sure the PACl is dripping into the Vertical T-junction.\n",
        "6. Open the ball valve directly below the Vertical T-junction.\n",
        "7. Volumetrically measure the flow rate of water through the reactor using the effluent tubing. This flow rate can be used to calculate the upflow velocity. The calculation is shown in the [Manual section](#Grav) of the report.\n",
        "8. Before stopping the experiment, collect an influent sample from the Fluoride Stock Tank and an effluent sample from the effluent tubing. These samples can have their fluoride concentrations measured using the fluoride probe and ProCoDA box available on the Automated system lab bench."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0kH2CNCtNHez",
        "colab_type": "text"
      },
      "source": [
        "### Automated System\n",
        "The Automated System uses the Process Control and Data Acquisition (ProCoDA) software to run experiments and log data. See the [Manual](#ProCoDA) for using ProCoDA to run experiments and gather data."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zB_2Q-qM39qE",
        "colab_type": "text"
      },
      "source": [
        "### Defining System Failure\n",
        "At the start of the summer, the team observed that, in the majority of experiments run, the floc blanket built up past the floc weir and sometimes gelled on the plate settlers before the floc weir was completely filled. The team defined this state of the system as failure, because it indicated that the floc weir was not adequately removing flocs from the sedimentation tube.\n",
        "\n",
        "In addition, gelling was a phenomenon that occurred sporadically in both red dye and fluoride experiments in Summer 2019 and previous semesters. An essential characteristic of a properly working floc blanket is continuous recirculation of flocs. Gels form when a large accumulation of flocs builds up so much that the flocs within it are unable to recirculate or move, and other separate flocs are forced to flow through a path around it rather than through it. The Summer 2019 team, like the Fall 2018 team, defined any buildup of gel, whether in the bottom of the reactor or on the plate settlers, to signal failure in the system.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/Failure.gif?raw=true\" height=\"400\">\n",
        "  \n",
        "  **Figure 18:** An unsuccessful red dye experiment. The floc blanket built up significantly on the plate settlers before the floc weir was filled.\n",
        "    </center>\n",
        "\n",
        "The importance of defining how a system fails is key to this research. Previous semesters had removed ionic fluoride, which was evident through the high uptake values seen in the data. However, the nature of the experiments were quite different from those run by the Summer 2019 team. The previous semester experiments ran for several hours and sometimes overnight, while in contrast the Summer 2019 experiments were run and monitored closely over a period of one to two hours. Previous teams did not focus on seeing if flocs were actually exiting through the floc weir either. Thus, it is possible that previous semesters were not removing flocs and were instead sending flocs through the effluent that were not detectable by the fluoride probe.\n",
        "\n",
        "  <center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/Working%20Floc%20Weir.gif?raw=true\" height =\"300\">\n",
        "  \n",
        "**Figure 19:**  A successful experiment, with the flocs exiting through the floc weir instead of building up on top of the plate settlers.\n",
        "    </center>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mk7OkZSMO87v",
        "colab_type": "text"
      },
      "source": [
        "###Adjusting PACl Concentration\n",
        "One goal of the fluoride subteam is to determine the ideal conditions for flocculation in the fluoride system. Because there is not a plethora of research done looking at the interactions between fluoride floc formation and polyaluminum chloride, this was one area of interest for the team. In particular, the team wanted to ensure that PACl would be added in amounts that would meet the WHO standard of 1.5 mg/L, but also in low enough concentrations so that there would not be an excessive need for coagulant in the system. Some plants that may use this technology could be remote, and thus supply chain and delivery of coagulant can be difficult to control. For this reason, the goal was to use as little coagulant as possible.\n",
        "\n",
        "In past years, teams used large PACl to fluoride proportions; sometimes as large as 8:1. While these tests yielded positive results with a large uptake of fluoride per unit of coagulant, they were also using concentrations of PACl in excess of 40 mg/L, which is relatively a large amount of coagulant. The team started with PACl to fluoride ratios in this range, and then reduced PACl concentrations in subsequent tests to determine if there was a point where the WHO standard was exceeded. For more details on the outcomes from these tests, please see the results & analysis section."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b__f-_7oofZC",
        "colab_type": "text"
      },
      "source": [
        "### Floc Blanket Height\n",
        "\n",
        "The height of the floc blanket over the course of each experiment was of interest because the team wanted to determine if there was a noticable trend in fluoride effluent with respect to the height of the floc blanket. If the fluoride effluent in the effluent decreased as the floc blanket height increased, a possible conclusion would be that the floc blanket does in fact contribute to the removal of dissolved fluoride. \n",
        "\n",
        "The floc blanket height was measured every five minutes from the top of the bottom junction on the sedimentation tube. A tape measurer was used to measure the height in centimeters, and the height was recorded in the text data logging interface in ProCoda. Afterwards, these data points were extrapolated from the text logging and entered as numeric data in Excel. For more on these data, please see the [Results](#res) section.\n",
        "\n",
        "In another effort to understand the implications of the height of the floc blanket, the team also increased the length of the sedimentation tube significantly. Keeping all other values constant, it was hypothesized that if there was any noticeable change in fluoride effluent with a longer sedimentation tube, this change could be attributed to the freshly increased floc blanket height. The layout of this new sedimentation tube can be seen above in Figure 14.\n",
        "\n",
        "An in-depth analysis of this configuration can be seen in the [Results](#res) section, however it was concluded that this configuration did not significantly improve effluent fluoride concentrations, and thus the design was deemed unnecessary."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xA8GflEobeGP",
        "colab_type": "text"
      },
      "source": [
        "### Jar Tests\n",
        "\n",
        "The Fluoride Summer 2019 team decided to test whether the removal of fluoride was better through a continuous or batch system. Therefore, the team performed two types of jar tests to analyze the behavior of fluoride and PACl when PACl is added all at once (40 mg/L) or in portions(4 doses of 10 mg/L). The purpose of the jar test was to see if different methods of adding PACl into the system affected the efficiency of the removal of fluoride. More details are available in the [results section](#jartest).\n",
        "\n",
        "Jar Test 1 Procedure (Batch):\n",
        "1. Make 1 liter of 5 mg/L fluoride solution in a jar.\n",
        "2. Measure the concentration of the solution using the fluoride probe. Keep the fluoride probe in the jar during the entire experiment.\n",
        "3. Add 140.6 μL of 71.12 g/L PACl to the fluoride solution.\n",
        "4. Stir for 10 seconds and let the solution sit for 7 minutes. Record the concentration after 7 minutes.\n",
        "4. Repeat step 3 and 4 three more times until 560 μL of 71.12 g/L PACl is added.\n",
        "\n",
        "Jar Test 2 Procedure (Continuous):\n",
        "1. Make 1 liter of 5 mg/L fluoride solution in a jar.\n",
        "2. Measure the concentration of the solution using the fluoride probe. Keep the fluoride probe in the jar during the entire experiment.\n",
        "3. Add 562 μL of 71.12 g/L PACl to the fluoride solution.\n",
        "4. Stir for 10 seconds and let the solution sit for 7 minutes. Record the concentration after 7 minutes."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_GN0525gV7wm",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Results_and_Analysis\"></a>\n",
        "## Results and Analysis"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gsl005kyesDi",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Consistent_Flocculation\"></a>\n",
        "### Consistent Flocculation in the Fluoride Gravity System\n",
        "\n",
        "The Fluoride Gravity Spring 2019 team's vertical T-junction system was last implemented in the Gravity System. The team observed successful flocculation of red dye and PACl in the flocculator but did not fully test if flocculation was consistent enough to create a fluidized bed in the sedimentation tube. Therefore, the Summer 2019 team ran initial experiments with red dye and PACl in the gravity-powered system to see if consistent flocculation would occur and if a fluidized bed would then build up to the height of the floc weir.\n",
        "\n",
        "The team observed consistent flocculation and a fluidized bed formation throughout the sedimentation tube. Subsequently, the team started testing fluoride with the gravity system. The gravity system has since been flocculating consistently with fluoride and PACl in the system, forming a fluidized bed, and having the flocs entering the floc weir (Figure 20). \n",
        "<center>\n",
        "<img src = \"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/grav%20increased%20angle%20success.gif?raw=true\" height =\"300\">\n",
        "\n",
        "**Figure 20:** The fluoride flocs built a fluidized bed up to the floc weir and accumulated in the floc weir.\n",
        "    </center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6eGgb5N4GXfj",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Upflow_Velocity\"></a>\n",
        "### Determining an Optimal Upflow Velocity\n",
        "Existing AguaClara plants use an upflow velocity of 1 mm/s for the sedimentation process. However, previous fluoride teams used an upflow velocity of 1.5 mm/s to graph points on the Langmuir Isotherm to prevent gelling in the tube. One of the goals from the Spring 2019 Automated team was to figure out if there was possibly an optimal upflow velocity between 1 mm/s and 1.5 mm/s that was low enough to ensure that the floc blanket could form properly but high enough to prevent gel from building up in the sedimentation tube. \n",
        "\n",
        "The Summer 2019 team started its experiments with an upflow velocity of 1 mm/s in order to mimic the existing AguaClara plants. However, after observing the flocs gelling on the plate settlers in many of the experiments, the team decided to try a higher upflow velocity to see if it would change the outcome of the experiment. Both with and without clay, the team ran qualitative experiments in the Automated and Gravity systems to see how the floc blanket would behave if the upflow velocity was higher than 1 mm/s.\n",
        "\n",
        "The water pump in the Automated system presented a slight complication when the team realized that influent samples of fluoride were much lower than expected (for example, several experiments that were supposed to have 5 mg/L of influent fluoride were measured to have an actual concentration of between 3 and 4 mg/L of fluoride). The team proceeded to troubleshoot the system, starting with the water pump. The output of the pump was measured volumetrically and the team determined that the number of mL of water pumped out for each revolution of the pump's rotors was not 0.8 mL/rev as indicated by the tubing size, but was instead 1 mL/rev. The implication of this discovery was that the tests believed to be run at 1 mm/s since the beginning of the summer had actually been running at around 1.32 mm/s. The team documented this change for future reference. Additionally, the Gravity system ran at a slightly slower 1.17 mm/s throughout the summer, with comparable results. Thus, the team believes that a range of upflow velocities can work for fluoride removal, and will continue to explore this idea in future semesters to determine a reasonable range."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jjsB1bzMe2a2",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"botgeo\"></a>\n",
        "### Choice of Bottom Geometry in the Sedimentation Tube\n",
        "Fluoride teams in 2016 began fabricating bottom geometry inserts for the bottom of the sedimentation tubes, where the influent water enters. This component of the apparatus was meant to mimic the jet reversers of the field plants, which allow for resuspension and recirculation of flocs during sedimentation. The goal of using the bottom geometry was to eliminate horizontal surfaces to prevent the settling of flocs and subsequent gelling or clogging of the sedimentation tube. Currently, the High Rate Sedimentation (HRS) Bottom Geometry team fabricates the bottom geometry inserts that the Fluoride teams use.\n",
        "\n",
        "In the Fluoride Gravity Spring 2019 team, the addition of a bottom geometry obtained from the [HRS Bottom Geometry Team](https://github.com/AguaClara/HRS-Bot-Geo) prevented further gelling in the sedimentation tube and allowed for a fluidized bed to form.\n",
        "\n",
        "<center>\n",
        "  \n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride_Gravity/master/Spring%202019/Botgeo%20Drawing.jpg\" height = 400 width = 500>\n",
        "\n",
        "**Figure 21:** 2D scaled drawing of the bottom geometry that was added to the Fluoride Gravity sedimentation tube.\n",
        "  \n",
        "  </center>\n",
        "\n",
        "On the other hand, the Fluoride Auto Spring 2019 team received a different bottom geometry obtained from [High Rate Sedimentation Team](https://github.com/AguaClara/HRS-Bot-Geo) for their sedimentation team as seen in Figure 18.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Images/BotGeo%20Drawing.jpg\" height = 400 width = 500>\n",
        "\n",
        "**Figure 22:** 2D scaled drawing of the bottom geometry used by the Spring 2019 Automated system that the Summer 2019 team did not end up using.\n",
        "  </center>\n",
        "  \n",
        "To compare the efficiency of the two bottom geometries, the Fluoride Summer 2019 team ran two experiments with the same dosages of red dye and coagulant as well as using the same sedimentation tube to observe which bottom geometry should be used for the sedimentation tube. The experiments were conducted on the Fluoride Automated system.\n",
        "\n",
        "While testing the bottom geometry from the Spring 2019 Fluoride Gravity apparatus, no gel formed in the sedimentation tube, and the floc blanket grew properly as a fluidized bed. However, when testing the bottom geometry from the Spring 2019 Fluoride Auto apparatus, gel appeared at the bottom of the sedimentation tube. Therefore, the optimal bottom geometry for the sedimentation tube for fluoride removal was determined to be design shown in Figure 21; this bottom geometry was used for all following experiments during Summer 2019."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OeRhQH_-a2Vh",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Troubleshooting_Gel\"></a>\n",
        "###Troubleshooting Gel in the System\n",
        "Gel occurred in both the Automated and Gravity system several times throughout the summer, even after the initial changes made to the bottom geometry and sedimentation tube angle. For example, on July 11 the team ran an experiment in the Automated system with flow rates that had not previously gelled, but gel built up throughout the tube and on the plate settlers. However, after making a fresh solution of PACl, the team stopped observing gel for the rest of the experiments that week. Thus, the team hypothesizes that the possible oxidation of PACl or some other interaction between PACl and species in the air may have caused the PACl to function less effectively as a coagulant. While the PACl Stock Tank was covered at all times to prevent evaporation, it was still not in a completely air tight chamber. The team will continue to investigate this hypothesis going forward.\n",
        "\n",
        "On other occasions, gel would be observed at the bottom of the sedimentation tube from the beginning of the formation of the floc blanket, and the team considered that the bottom geometry was somehow clogged. Several times after running high-velocity water from the sink through the bottom geometry the system would stop gelling, which indicates that a congested bottom geometry or influent tube to the sedimentation tube may also prevent the formation of a proper floc blanket or cause a gel from the beginning of the sedimentation process.\n",
        "\n",
        "Finally, the team had used a lower PACl concentration in the Gravity system compared to the Automated system for the majority of the summer. A reactor concentration of 18.9 mg/L of PACl was used in the Gravity system, while 30-40 mg/L of PACl was used in the Automated system. When the team tried to implement a higher concentration of 30.1 mg/L of PACl in the Gravity system in order to reach a closer concentration to the one used in the Automated system, gelling was observed. Another variable was the upflow velocity; the Gravity system had been running at around 1.17 mm/s while the Automated system had been running at around 1.32 mm/s. The team believed that the upflow velocity in the Gravity system was too low and possibly facilitating the formation of gel at the higher PACl concentration, so the upflow velocity was increased to around 1.55 mm/s (the team still needs to improve the method of adjusting heights and upflow velocities with more precision in the Gravity system). However, despite this higher upflow velocity in the Gravity system, gel was still observed from the beginning of the floc blanket formation. Thus, the team concluded that the Gravity system is more sensitive to PACl dosages, and aims to understand in future semesters why the same PACl concentration in the Automated reactor behaves differently from the Gravity reactor."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mEn3SsReWee6",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Calculations\"></a>\n",
        "### Sedimentation Tube Experimental Calculations\n",
        "The following equation, from the Fall 2017 High Rate Sedimentation team report [(Galantino et. al)](https://github.com/AguaClara/high_rate_sedimentation/blob/master/Reports/HRS%202017%20Fall%20Final%20Report.pdf), determines the capture velocity of a system given the inner diameter of the sedimentation tube (S), the upflow velocity (V<sub>a</sub>), and the length of the plate settler after the floc weir (L). \n",
        "<center>\n",
        "\\[V_{c}=\\frac{SV_{a}sina}{Lsinacosa+S}\\]\n",
        "</center>\n",
        "\n",
        "Using this equation, the Summer 2019 team was able to use the system parameters of a 1\"-wide tube, an upflow velocity of 1.32 mm/s in the Automated system, and a length L of 23.5 cm to determine the capture velocity of the system to be 0.312 mm/s. In the Gravity system with an upflow velocity of 1.17 mm/s, the capture velocity is 0.277 mm/s.\n",
        "\n",
        "This capture velocity may change in the future if the team decides to lengthen the plate settler portion at the top of the sedimentation tube or make other adjustments to the design in an attempt to decrease the amount of fluoride escaping through the effluent."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vsIBMlzoWB6X",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Sed_Tube_Design\"></a>\n",
        "### Sedimentation Tube Design\n",
        "\n",
        "The Summer 2019 team tested the sedimentation tube from the Spring 2019 Automated team, in which the floc weir was located on the left side before the plate settler using red dye and PACl and observed the red dye flocs passing the floc weir and building up on the plate settler rather than entering  the floc weir (Figure 23). The flocs on the plate settler then formed a gel at the bend since the flocs failed to slide off and recirculate. When the team turned on the waste pump and opened the waste valve, the flocs started entering the floc weir. Therefore, the sedimentation tube with the floc weir on the vertical portion of the tube did not collect flocs in the floc weir when the valve was closed, and required the waste pump as a force to pull the flocs out.\n",
        "\n",
        "<center>\n",
        "  \n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Auto%20Weir%20Not%20Working%201.jpg\" height = 420 width = 260>\n",
        " \n",
        "\n",
        "**Figure 23:**  Gel building on the plate settler of the Spring 2019 Automated  sedimentation tube when the waste valve was closed.\n",
        "  </center>\n",
        "\n",
        "The Summer 2019 team then tested Spring 2019 Gravity team sedimentation tube with the floc weir on the plate settler using the same dosages of red dye and PACl as the previous experiment and observed that flocs formed a gel on the plate settler and did not reach the floc weir (Figure 24).\n",
        "\n",
        "<center>\n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Gel%20Grav%201.jpg\" height =420 width = 260>\n",
        "\n",
        "**Figure 24**: A gel forming on the plate settler of Spring 2019 Fluoride Gravity sedimentation tube when the waste valve was closed.\n",
        "  </center>\n",
        "\n",
        "In an effort to prevent the formation of a gel on the plate settler, the Summer 2019 team increased the angle of the plate settler to the vertical section of the sedimentation tube from 150° to 160° and observed that flocs entered the floc weir consistently without gel formation on the plate settler when using the design shown in Figure 25. As a result, the optimal sedimentation tube design was determined to be the sedimentation tube with the floc weir on the plate settler. The Summer 2019 team fabricated another 160° sedimentation tube to run tests on both the Gravity apparatus and Automated apparatus throughout the summer.\n",
        "\n",
        "<center>\n",
        "  \n",
        "<img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Increased%20Angle%20Success.jpg\" height = 420 width = 260>\n",
        "\n",
        "**Figure 25**: The flocs accumulating in the floc weir and no gel formation on the plate settler with Fluoride Gravity Spring 2019's sedimentation tube.\n",
        "  \n",
        "  </center>\n",
        "  \n",
        "  \n",
        "  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IM5UQlV51WfH",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Introducing_Clay\"></a>\n",
        "### Introducing Clay into the System\n",
        "A few weeks into the summer semester, the team began to think of a different approach of removing fluoride. As aforementioned, fluoride in both its ionic and floc form is nearly invisible and in the lab requires a flashlight held against the sedimentation tube in order to view the floc blanket. One hypothesis was that clay would both increase the turbidity of the flocs and possibly help the flocs settle rather than exiting the tube with the clean effluent water.\n",
        "\n",
        "The team decided to try adding clay into the system, with the goal of increasing the visibility of the fluoride flocs to both the turbidimeter and to the naked eye. Additionally, since clay particles are more massive than fluoride particles, the team hypothesized that the addition of clay would increase the mass of the flocs and thus allow the flocs to fall into the floc weir more easily rather than continuing to build up past the floc weir as they had been in experiments conducted before the addition of clay. The clay pump was added after the flocculator so that the fluoride would have sufficient time to bind with the PACl in the flocculator and not have to compete with the clay particles. Because of the placement of the pump, the team set a timer for 5 minutes after starting each experiment before turning on the clay pump, to give the fluoride and PACl time to make it through the flocculator before adding clay. The residence time of the flocculator, when using an upflow velocity of 1.3 mm/s as the team had been doing, is actually around 1.9 minutes, so the first flocs entered the tube as just fluoride and PACl and no clay.\n",
        "\n",
        "The addition of clay to the fluoride system was not a completely new change to the AguaClara fluoride removal research. The Spring 2016 team added clay to their system but decided that it was not a necessary component because it caused the effluent turbidity to rise and did not aid in the removal of fluoride, by their standards. \n",
        "\n",
        "The previous section details the attempts to understand the fluoride probe and ensure that the effluent fluoride is in a form that the probe can sense. However, the point of adding the clay was to try a different way of gauging success, by tracking turbidity rather than effluent fluoride concentration. The team observed an increase in turbidity while cleaning out the system and sending a large number of flocs through the effluent, but a significant increase in turbidity was not seen during regular experiments even when flocs were seen exiting through the effluent. At this point in the semester, the team is still unsure of the effect that the flocs have on turbidity. In comparison, the previous fluoride teams observed a much more definite and concrete increase in effluent turbidity when adding clay to the system. Thus, the Summer 2019 hypothesis was that the clay would bind to the fluoro-aluminum complexes and make the nearly clear flocs more turbid, which could then be used as an indicator for failure or success. An increase in turbidity could mean that fluoride was also escaping and would thus signal failure.\n",
        "\n",
        "The team compared experiments done with and without clay using the same dosages of PACl and fluoride and observed that using clay did not significantly increase the fluoride uptake per amount of PACl used as seen in Figure 15. The experiments done without clay either did similarly or better than the experiments with clay (Figure 26).\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Summer%202019%20Graph.png\">\n",
        "  \n",
        "**Figure 26:** The data from the Summer 2019 semester plotted with the clay experiments distinguished from the non-clay experiments.\n",
        "    </center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PY-pPGcoh_Ul",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Floc_Weir\"></a>\n",
        "### Increasing the Size of the Floc Weir\n",
        "The team wanted to find another way to capture more flocs by altering the design of the sedimentation tube. One hypothesis was that the size of the floc weir entrance was too small for large groups of flocs to pass through. The fluoride sedimentation tubes since Fall 2016 all used 1\" inner diameter PVC for the sedimentation tube and plate settler and 3/4\" inner diameter PVC for the floc weir. However, the choice to use a thinner tube for the floc weir was originally an arbitrary fabrication decision, so the team decided to see how a larger weir would affect the removal of flocs.\n",
        "\n",
        "The team fabricated a new sedimentation tube using the same tubing lengths as the previous sedimentation tubes. Instead of welding the floc weir, the team used PVC cement to connect the top and bottom of the sedimentation tube to a tee, and attached another length of 1\" inner diameter PVC tubing to the middle of the tee. This sedimentation tube was tested with fluoride and PACl but without clay. The team observed gelling at the entrance of the floc weir as seen in Figure 27.\n",
        "  \n",
        "  <center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/Large%20Weir%20(better%20quality).gif?raw=true\" height =\"300\">\n",
        "  \n",
        "**Figure 27:** The initial experiment in the sedimentation tube with the large floc weir. Flocs built up at the entrance of the floc weir and seemed to stop moving while flocs on the left side were able to move around it.\n",
        "    </center>\n",
        " \n",
        "The team later bent the tee to increase the angle between the plate settler and the floc weir and thus increase the force of gravity acting on flocs sliding down the weir, with the hope that the flocs would move faster down the weir instead of gelling at the entrance and preventing other flocs from entering.\n",
        "However, human error during the bending of the tee caused a lip to appear, which required more PVC glue to make the sedimentation tube water tight. The team ran further tests and a gel formed on the bottom corner of the T-junction, but still allowed the flocs to fall in the weir. The team hypothesized that the tee was not bent perfectly, and instead introduced a flat surface on which flocs could settle and eventually form a gel. Therefore, the team may in the future use a wye to fabricate a sedimentation tube in order to eliminate flat surfaces and make the fabrication of sedimentation tubes more uniform. \n",
        "\n",
        "The sedimentation tube with the larger floc weir was compared with the previous sedimentation tubes with the smaller floc weir by running experiments with identical dosages of PACl and fluoride in both the Automated and Gravity systems. The sedimentation tube with a larger floc weir did not produce results that were significantly different from those observed in the original narrower floc weir."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GZyt65AAIDwZ",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Fluoride_Probe\"></a>\n",
        "\n",
        "### Testing the Limits of the Fluoride Probe\n",
        "Effluent water from a fluoride experiment may contain both fluoride ions that were unable to flocculate and fluoride in the form of flocs, which are complexes of aluminum-hydroxyl and fluoride. Ion selective electrodes are designed to measure the concentration of ions in solution; thus, flocs cannot always be detected by the probe. The Spring 2019 team had collected data that showed an increase in the fluoride concentration reading after filtering through mesh, so the Summer 2019 repeated the same type of tests by taking samples of fluoride from the flocculator and measuring their concentrations before and after filtering. However, these new tests proved to be inconclusive, as many of the samples showed an insignificant increase in the concentration reading following filtration. The team is still unsure of why the results are so different from those in Spring 2019; one possible reason is that the probe used in Spring 2019 was nearly at the end of its life cycle and may not have been as accurate as it could have been, whereas the Summer 2019 team purchased a new probe at the beginning of June.\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Filtered%20Samples.png\" height=\"300\">\n",
        "  \n",
        "**Figure 28:**  Data collected from before and after filtering floc-filled samples through mesh.\n",
        "    </center>\n",
        " \n",
        "A study by Sharma et. al. shows the dependence on the flocculation of fluoride or decomplexing of flocs on the pH of a system. The study used a pH of 6.5-8.5 for \"minimum solubility of aluminium\" ([2015](https://www.ijarse.com/images/fullpdf/1441435538_393P.pdf)). Since pH is an integral aspect of flocculation that had not been previously discussed in depth in other semesters of research on fluoride removal, the Summer 2019 sought more information on it. The comprehensive manual provided by the probe manufacturer suggested using Total Ionic Strength Adjustment Buffer (TISAB) I or III as a way to create the correct pH in the solution at which fluoro-aluminum complexes would be able to break apart.\n",
        "\n",
        "Differing sources disagreed over the use of the different buffer solutions (TISAB I, II, and III). Since the AguaClara lab had access to TISAB II, the fluoride team tried measuring several floc-filled samples before and after adding TISAB II in a 1:1 ratio to the existing solutions. However, these tests also proved to be inconclusive, as the results continued to vary (many of the sample concentration readings went down after adding TISAB, even when considering the added volume from the TISAB).\n",
        "\n",
        "Later, longer term experiments showed a spike in turbidity but a dip in fluoride concentration when flocs were seen exiting the sedimentation tube. This observation suggested that the probe is unable to detect fluoride in the form of flocs. In addition, flocs were observed floating in several effluent samples, which indicates that flocs were not being filtered out by the mesh either."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sk-fHGAc3lWZ",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Fluoride_Turbidity\"></a>\n",
        "### Fluoride Concentration and Turbidity over Time\n",
        "The team also ran several longer term experiments to see how fluoride concentration and turbidity of the system changed over time. They hypothesized that a high measured fluoride concentration would indicate that ionic, unflocculated fluoride was leaving the system, while a high turbidity would indicate that fluoride flocs were leaving the system. \n",
        "\n",
        "Figure 29 below displays fluoride concentration and turbidity data taken during the course of an experiment, in which steady state was reached in less than one hour. In several similar experiments, a dip in turbidity corresponded with an increase in fluoride concentration, as seen around the 20 minute mark. This observation is consistent with the hypothesis that rising fluoride concentration indicates more ionic fluoride in the effluent and subsequently fewer flocs leaving the sedimentation tube and a lower turbidity.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/7-12-19%20Effluent%20Fluoride%20Concentration%20and%20Turbidity.png?raw=true\">\n",
        "  \n",
        "  **Figure 29:** Fluoride experiment run in the Automated system with an upflow velocity of 1.32 mm/s, 5 mg/L fluoride, and 35 mg/L PACl.\n",
        "  </center>\n",
        "  \n",
        "  Additionally, these tests showed that, for an upflow velocity of 1.32 mm/s, the steady-state concentration of fluoride occurred in less than or by one hour into the experiment.\n",
        "  \n",
        "  Finally, an initial valley in the fluoride concentration graph appeared for several of the experiments, indicating that it may be significant. However, the team has still not figured out why this dip in fluoride concentration occurs early on in the experiment.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fahsJHLGL8Se",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Blanket_Concentration\"></a>\n",
        "### Floc Blanket Height and Fluoride Concentration <a name=\"res\"></a>\n",
        "One concept that the team looked into was the relationship between the height of the floc blanket and the effluent fluoride. Theoretically, a decreasing fluoride concentration with an increasing floc blanket height would indicate that the height of the floc blanket matters in the design of the sedimentation tube, in order to optimize uptake of fluoride by PACl. It would warrant a redesign of the sedimentation tube in order for the floc blanket to occupy more vertical space.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-10-19/710%20f%20conc.png\">\n",
        "  \n",
        "**Figure 30**: Fluoride concentration over time on 7/10/2019. Test run at 40 mg/L PACl and 5 mg/L fluoride, with an upflow velocity of 1.3 mm/s.\n",
        "</center>\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-10-19/710%20blanket.png\">\n",
        "  \n",
        "**Figure 31**: Height of the fluoride floc blanket over time on 7/10/2019. Test run at 40 mg/L PACl and 5 mg/L fluoride.\n",
        "</center>\n",
        "\n",
        "\n",
        "Figures 31 and 31 above show how the height of the floc blanket and the effluent fluoride concentration changed over the course of 33 minutes on a particular test. Although this provided some insight and showed that fluoride concentration seemed to decrease as the floc blanket grew, the team considered that this may be attributed to the fact that the test typically takes longer (around 1 hour 30 minutes) to reach a steady state condition. For this reason, the team decided to lengthen the sedimentation tube, as shown in Figure 14 in the Methods section, to see if a drastically increased floc blanket height would have any effect on the effluent fluoride concentration. In this case, the team effectively doubled the length of the floc blanket. \n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-24-2019/724%20both.png\">\n",
        "  \n",
        "**Figure 32**: Height of the fluoride floc blanket over time for 7/24/2019. Test run at 40 mg/L PACl and 5 mg/L fluoride.\n",
        "</center>\n",
        "\n",
        "Figure 32 above shows the data recorded for the experiment run with the lengthened sedimentation tube. The team noticed no correlation between floc blanket height and effluent fluoride concentration (at a certain point, the fluoride concentration would level out instead of continuing to decrease even as the floc blanket grew taller). Additionally, it was observed that using a longer sedimentation tube did not produce significantly lower effluent fluoride concentrations than the original sedimentation tube. Tests run with the same concentrations of fluoride and PACl and the same upflow velocity in the original sedimentation tubes resulted in steady-state fluoride effluent levels around 1.3 mg/L, on average. The test that used the long sedimentation tube resulted in a steady-state fluoride concentration of approximately 1.2 mg/L. It was determined that this decrease was not a significant enough improvement to justify using the longer sedimentation tube."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4ymu3Y9TapHE",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Horizontal_Reactor\"></a>\n",
        "###Testing a Horizontal Reactor\n",
        "One relatively new area of AguaClara technology is a horizontal sedimentation tube ([Figure 15](#Figure_15)<a href=\"Figure_15\"></a>) as opposed to a vertical sedimentation tube. The main purpose of a horizontal reactor is to avoid the formation of a floc blanket, and to allow clean water to travel upwards towards the effluent tube while the flocs continue traveling horizontally to a waste tube. The water in the effluent tube must run at a higher velocity than the water through the main tube, so it is necessary to attach another pump to the effluent tube. The team used a horizontal reactor from a previous team to test a water velocity of 1 mm/s, 2 mm/s, and 5 mm/s. The team also decreased the PACl concentration for each increase in water velocity (due to limitations of the current coagulant pump). However, since the team had never used a horizontal reactor in previous years, the observations were mostly qualitative.\n",
        "\n",
        "An accumulation of fluoride flocs and gel was seen in the sedimentation tube directly next to the influent valve, even for the experiments with water velocities as high as 5 mm/s. The team hypothesized that a higher flow rate of influent water was needed in order to prevent the flocs from settling in the reactor. However, the team deemed the horizontal reactor less feasible for testing because of the large amounts of wasted water compared to the vertical reactors and the high water velocities required. Thus, the team went back to using the vertical sedimentation tube setup.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Horizontal%20Reactor%20Gel.gif\" height=\"200\">\n",
        "  \n",
        "  **Figure 33:** Gel forming at the entrance of the horizontal reactor and impeding the flow of flocs through the tube.\n",
        "</center>\n",
        "  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_Nn9zXKc-Fjq",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Isotherm\"></a>\n",
        "###Fitting the Experimental Data to the Isotherm\n",
        "The Summer 2018, Fall 2018, and Spring 2019 teams ran experiments with different influent ratios of PACl to fluoride, and the data was plotted by the Spring 2019 team. The Summer 2019 team ran the experiments in the following [document](https://docs.google.com/document/d/1892_u8ZQFAalcHQAYuL7N48HOVAAVtQmxul6kjME7j8/edit?usp=sharing) to see if the new data would match the trends of the existing data. However, the Summer 2019 data points all appeared much lower than both the best-fit line from previous semesters and the theoretical Langmuir Isotherm developed by the Spring 2019 team. As shown in the graph below, the experiments run during the Summer 2019 resulted in lower uptakes than those achieved in previous semesters, which would indicate a lower efficiency of PACl in removing fluoride than in prior semesters.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Cumulative%20Graph.png\" >\n",
        "\n",
        "**Figure 34:** The data from the past three semesters and the current team plotted against the theoretical Langmuir Isotherm.\n",
        "  </center>\n",
        "\n",
        "The graph below emphasizes the contrast between the best-fit line for the Summer 2019 data and the Langmuir Isotherm developed by previous semesters.\n",
        "  \n",
        "  <center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Summer%202019%20Data%20with%20Previous%20Isotherm.png\">\n",
        "  \n",
        "  **Figure 35:** The Summer 2019 data plotted against the theoretical Langmuir Isotherm developed by previous Fluoride teams.\n",
        "  \n",
        "  R2 value for best-fit line: 0.7923313195369291\n",
        "\n",
        "  </center>\n",
        "    \n",
        "The discrepancy in the graph and the differences in the apparatus between Summer 2019 and previous semesters - namely, the angle of the sedimentation tube and the lower upflow velocity - prompted the Summer 2019 team to develop its own Langmuir Isotherm with different constants that would better model the data collected this semester. This model is shown in Figure 35 below.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Summer%202019%20Data%20with%20Current%20Isotherm.png\">\n",
        "  \n",
        "  **Figure 36:** Theoretical Langmuir Isotherm based off of Summer 2019 data only.\n",
        "    Q = 555.7, K = 0.12\n",
        "  </center>\n",
        "  \n",
        "  The results indicate that the Langmuir Isotherm is still a feasible model for describing the interaction between PACl and ionic fluoride. However, the R2 value is still quite low and there has not yet been confirmation about the nature of the bonds that form between PACl and fluoride, so the team is open to the consideration that another model may more reasonably fit the data."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H1QKyqKGgSM0",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Molecular\"></a>\n",
        "### Molecular Interactions Between PACl and Fluoride\n",
        "The Summer 2019 team consistently used tests with 5 mg/L of fluoride and 30-40 mg/L of PACl. Compared to AguaClara teams that work with other contaminants such as clay, this dosage of PACl is quite high. Further, many of the experiments even showed that these dosages were sometimes not sufficient to remove fluoride in order to meet the WHO standard, meaning that it is possible that a greater concentration of PACl is necessary.\n",
        "\n",
        "The team wanted to see how many fluoride atoms were actually binding to PACl molecules. With a molecular formula of Al2(OH)3Cl3, PACl has a molar mass of 211.5 g/mol. Using this molar mass and the molar mass of ionic fluoride (19.0 g/mol), the number of molecules of each species being put into the system was calculated. The number of atoms of fluoride leaving in the effluent was also computed, and the difference between the influent and effluent was assumed to be the atoms of fluoride in floc form, since the team established that the probe cannot detect flocs, and because mass must be conserved. \n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Molecular%20Interactions%20PACl%20and%20Fluoride.png\" height=400>\n",
        "  \n",
        "  **Figure 37:** Table showing the number of atoms of fluoride removed per molecule of influent PACl for trials with different influent concentrations of each chemical. \n",
        "  </center>\n",
        "\n",
        "Currently, the team believes that the PACl may dissociate into Al(OH)3 and AlCl3 when added to the influent water. The Al(OH)3 may then bond with the fluoride, possibly in a sort of net around the fluoride. The results of the table show that, the lower the ratio of fluoride to PACl, the more molecules of PACl there are for each atom of fluoride. While these results provide some insight into how fluoride and PACl bind and how their concentrations affect bonding, much more work must be done in order to confirm these calculations and figure out a way to use the information for better removal of fluoride in the future.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Uptakes%20of%20Fluoride%20Atoms%20by%20PACl%20Molecules.png\">\n",
        "  \n",
        "  **Figure 38:** Graph showing the number of fluoride atoms removed from solution per PACl molecule.\n",
        "  \n",
        "  </center>\n",
        "  \n",
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s2x8IqYEAiPE",
        "colab_type": "text"
      },
      "source": [
        "### Jar Test <a name=\"jartest\"></a>\n",
        "\n",
        "A closed system was tested to observe how flocs and fluoride concentration would behave in a closed batch system. A jar configuration was set up as described in the Procedure section, and then different concentrations of PACl were added to a 5 mg/L solution of fluoride. The most notable observations occurred when PACl was added and stirred into the solution for 10 seconds only, and then allowed to sit for 7 minutes, with the fluoride probe measuring continuously. It was observed that the change in fluoride concentration happened in the first few seconds when the solution was being mixed, and then the concentration leveled out for the remainder of the period before measurement. This can be seen by the behavior of the fluoride probe, shown below in Figure 38. This graph shows two additions of PACl, both occurring at the spikes. After these spikes, the voltage levels out in a few seconds. This signifies the short mixing period and subsequent period of settling (note that the higher the voltage on the probe, the lower the fluoride concentration). Based on this data, it was hypothesized that the current length of the flocculator and residence time (around two minutes with an upflow velocity of 1.32 mm/s) may be superfluous. Based on the jar test, the same removal efficiencies can be achieved after only a few seconds of mixing. Figure 39 below shows the actual concentration changing as a function of PACl dosage, with Table 1 providing the corresponding data used to create Figure 39. In reality, all of the fluoride removal happens within 15 seconds of adding the PACl.\n",
        "\n",
        "An instantaneous decrease in fluoride concentration as seen in Figure 39 may suggest that the entire mechanism for fluoride removal may happen in the flocculator (since comparable final fluoride concentrations were observed during the jar test and the sedimentation tube tests). Nevertheless, the instantatneous decrease in fluoride concentration may also indicate that the period of flocculation likely does not need to be as long as it is designed for now. According to Figure 35, the entire flocculation process could take place in the order of 10-15 seconds and adequate removal of fluoride would still be achieved.\n",
        " \n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/jartest.png\">\n",
        "  \n",
        "**Figure 39:** ProCoda data from probe measurements on diffusion test. Spikes indicate 40 mg/L injections of PACl.\n",
        "</center>\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-24-2019/jartest_better.png\">\n",
        "  \n",
        "**Figure 40:** Fluoride concentration change and corresponding PACl dose for the diffusion test.\n",
        "</center>\n",
        "\n",
        "Based on these tests, it was also confirmed that the fluoride probe does not recognize fluoride in the floc form. This was determined based on the fact that no fluoride left the system, yet a decrease in fluoride concentration was observed as more PACl was added. Table 1 below shows how the fluoride concentration changed as more PACl is added into the system. It was suggested that PACl and fluoride exist in a type of equilibrium where the first dose of PACl adsorbs a large amount of fluoride, but as more PACl is added, some fluoride will break away from the initial flocs to interact with the newly introduced PACl. This was further investigated by comparing data between two trials: the first where PACl was added in 10 mg/L injections and allowed to stir and sit until a total of 50 mg/L was added to the system, and also a dose of 40 mg/L up to 200 mg/L total. Table 1 below shows the former procedure data, and Table 2 the latter.\n",
        "\n",
        "<center>\n",
        "\n",
        "**Table 1:** Diffusion test with 40 mg/L injection of PACl into a 4.2 mg/L solution of fluoride.\n",
        "\n",
        "| Dose \t| Fluoride Concentration (mg/L) \t| PACl Dose (mg/L) \t| Percent Decrease (%) \t|\n",
        "|:----:\t|:-----------------------------:\t|:----------------:\t|:--------------------:\t|\n",
        "|   0  \t|          4.248738884          \t|         0        \t|           -          \t|\n",
        "|   1  \t|          1.230121699          \t|       39.8       \t|         71.0         \t|\n",
        "|   2  \t|          0.151319939          \t|       79.6       \t|         87.7         \t|\n",
        "|   3  \t|          0.012275881          \t|       119.3      \t|         91.9         \t|\n",
        "|   4  \t|          0.001184049          \t|       159.0      \t|         90.4         \t|\n",
        "|   5  \t|          0.000522253          \t|       198.6      \t|         55.9         \t|\n",
        "\n",
        "\n",
        "**Table 2:** Diffusion test with 10 mg/L injection of PACl into a 4 mg/L solution of fluoride.\n",
        "\n",
        "| Dose \t| Fluoride Concentration (mg/L) \t| PACl Dose (mg/L) \t| Percent Decrease (%) \t|\n",
        "|:----:\t|:-----------------------------:\t|:----------------:\t|:--------------------:\t|\n",
        "|   0  \t|          4.052880929          \t|         0        \t|           -          \t|\n",
        "|   1  \t|          3.198660453          \t|       9.96       \t|         21.1         \t|\n",
        "|   2  \t|          2.237519271          \t|       19.91      \t|         30.0         \t|\n",
        "|   3  \t|          1.515152036          \t|       29.86      \t|         32.3         \t|\n",
        "|   4  \t|          1.040381089          \t|       39.80      \t|         31.3         \t|\n",
        "|   5  \t|          0.731150354          \t|    49.74917558   \t|         29.7         \t|\n",
        "\n",
        "</center>\n",
        "\n",
        "\n",
        "Despite the different additions of PACl into the jars, these two tests yielded similar results. Therefore, the addition of PACl in several additions does not result in more successful fluoride removal than adding the PACl dosage all at once. Another notable observation was that, as more PACl was added past the 40 mg/L dosage, the fluoride concentration continued to decrease. This data could be useful in cases where increasing levels of fluoride are observed and higher levels of PACl may be needed to bring levels below the WHO standard."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1IAR3V5jgRgf",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Continuous_System\"></a>\n",
        "### Implementing the Jar Test in a Continuous System\n",
        "Although the jar test indicated that adding PACl in small doses did not increase performance, the team hypothesized that the floc blanket in a continuous system might have an effect on the performance of the PACl. Thus, the team converted the jar test setup to a continuous system. As seen in Figure 16 and Figure 17, two sedimentation tubes were connected in series, with a PACl injection between each tube. Initially, there was only a flocculator before the first sedimentation tube, while the tubing between the sedimentation tubes acted as a contact chamber for the flocs to form by diffusion. However, the flocs formed through diffusion appeared different from the flocs formed by flocculation; the diffusion flocs seemed to be less well defined in shape and did not fall down the floc weir like the flocculated ones did. The fluoride removal was also worse without flocculation before the second sedimentation tube, as shown in the graph below in Figure 40, where the steady state fluoride concentration is well above 2.0 mg/L and thus higher than the WHO standard. \n",
        "\n",
        "<center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/7-30-19%20Effluent%20Fluoride%20Concentration%20and%20Turbidity.png?raw=true\">\n",
        "  \n",
        "  **Figure 41:** Experiment run on 7/30/19 with two sedimentation tubes and a flocculator preceding the first sedimentation tube. 20 mg/L of PACl was added before each sedimentation tube for a total of 40 mg/L of PACl in the entire system.\n",
        "  </center>\n",
        "  \n",
        "  Thus, a flocculator was added after the first sedimentation tube and before the second. The experiments were then run with 5 mg/L of fluoride and 20 mg/L of fluoride before each sedimentation tube (or 40 mg/L of fluoride in total).\n",
        "\n",
        "Since the setup was fairly new (reactors in series were used briefly in 2016 but had inconclusive results), the team was unsure of when to begin adding the PACl to the second sedimentation tube. Thus, it was arbitrarily added when the floc blanket in the first sedimentation tube reached the height of the weir.\n",
        "\n",
        "A floc blanket was observed forming in both sedimentation tubes. The uptakes and final fluoride concentrations, however, did not seem to be an improvement from the system with the single sedimentation tube. Results are summarized in the table below.\n",
        "\n",
        "| Dose 1 (mg/L PACl) | Dose 2 (mg/L PACl) \t| Influent Fluoride (mg/L) |Effluent Fluoride  (mg/L)\t| Uptake \t|\n",
        "|:----:\t|:-----------------------------:\t|:----------------:\t|:--------------------:\t|:------------------------:|\n",
        "|   40  \t|          20          \t|         4.254284273        \t| Around 0.9|          55.90473788          \t|\n",
        "|   40  \t|          20          \t|       3.295733383       \t|  Around 0.8|       41.59555638        \t|\n",
        "|   20  \t|          20          \t|        3.42950010504444    \t|     Around 1.5  |  48.23750263         \t|\n",
        "\n",
        "Since the uptakes were relatively low compared to the single sedimentation tube system, the team decided not to continue using two sedimentation tubes in series."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Me12TWjY-Iys",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Conclusions\"></a>\n",
        "## Conclusions\n",
        "The Summer 2019 team determined a geometry for the sedimentation tube, with a 160° angle and the floc weir on the plate settler, for which fluoride flocs were seen exiting through the weir. While this single sedimentation tube system did succeed in removing flocs from the water, the effluent fluoride concentration was still above the WHO standard of 1.5 mg/L for many of the experiments. \n",
        "\n",
        "In an attempt to remove more fluoride flocs from the system, the team tried different variations of the system design. The first change was to add clay to the Automated system. This addition of clay did not significantly increase the measured uptake of fluoride by PACl in the system, so the team decided not to move forward with the clay system.\n",
        "\n",
        "The next variation of the system used a wider floc weir than the ones in previous sedimentation tubes. Fabrication errors caused a horizontal surface to remain inside the tee, which the floc weir was attached to, and cause the eventual buildup of gel at the tee even when it originally worked at removing flocs. The uptakes were calculated for this sedimentation tube before failure but did not display improvement from the performance of the sedimentation tube design with the narrower floc weir.\n",
        "\n",
        "Throughout the summer, the team wanted to figure out how the fluoride ion selective electrode (ISE) probe works, and if it detects fluoride in the form of flocs or just in ionic form. The jar tests and experiments that logged fluoride concentration and turbidity over time indicated that the probe does not detect fluoride in floc form. Further research into the probe revealed a possible need for a Total Ionic Strength Adjustment Buffer to decomplex fluoride-PACl flocs.\n",
        "\n",
        "Long-term experiments that were allowed to run to steady-state revealed an inverse relationship between effluent fluoride and effluent turbidity, which also further supports the claim that the fluoride probe cannot detect flocs.\n",
        "\n",
        "Fabricating a taller sedimentation tube with the ability to grow a taller floc blanket proved to be unnecessary, as the uptakes were not significantly higher than the ones observed with the original-sized sedimentation tube. This test showed that the floc blanket height does not seem to have a correlation with the amount of fluoride removed, and suggests that perhaps the floc blanket does not have to be as tall or even exist in the first place for fluoride removal to be possible.\n",
        "\n",
        "The horizontal reactor, which is meant to not utilize a floc blanket, was tried and abandoned after the repeated formation of gel at the entrance of the sedimentation tube.\n",
        "\n",
        "From the jar tests, the team saw that the binding of PACl to flocs happens in the first few seconds of contact, which may indicate that the length of the flocculator is not as important as it was originally believed to be. Additionally, the jar tests showed that adding the same amount of PACl in several small dosages is not more effective than adding all the PACl to the fluoride solution at once. \n",
        "\n",
        "Using a lab bench with two sedimentation tubes in series confirmed that adding PACl in multiple doses does not increase uptake of fluoride ions. These experiments also showed that the flocculator is necessary in order to form flocs that can properly fall down the floc weir.\n",
        "\n",
        "After plotting the Summer 2019 data and developing a Langmuir Isotherm, it was determined that the model is still feasible for describing coagulant-fluoride interactions but still requires further research and possibly a re-design of the system or the use of a different model in order to increase uptakes and meet the WHO standards."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uQBlaQNP-VJA",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Future_Work\"></a>\n",
        "## Future Work\n",
        "Going forward, the team plans to continue systematically testing different PACl-to-fluoride ratios in order to obtain an optimal coagulant dosage that will lead to an effluent concentration below 1.0 mg/L of fluoride. The influent fluoride concentrations to be tested will be between 1 and 10 mg/L, in order to simulate realistic groundwater conditions ([Rao et.al.](https://link.springer.com/article/10.1007/s00254-003-0873-3))\n",
        "\n",
        "Additionally, the team will continue trying different methods of decomplexing the aluminum hydroxyl fluoride complexes that make up flocs. The goal is to ensure that all of the fluoride in the effluent is in the form of ions rather than flocs so that all the fluoride can be detected by the ion selective probe. Once there is a reliable way to consistently decomplex flocs, the team will use a uniform measuring process for experiments going forward.\n",
        "\n",
        "Since it was found that more PACl is possibly needed to treat the influent concentrations of fluoride used during Summer 2019, the team is also beginning to consider the standards for acceptable levels of aluminum in drinking water, since some aluminum from the PACl does escape through the effluent. The EPA-mandated maximum aluminum concentration is 0.2 mg/L. Enough PACl must be dosed in the system to bring the fluoride concentration down to a healthy level, but there cannot be so much PACl that the aluminum concentration rises to an unsafe level. The study by Dubey et al. shows that a filter of 0.2 um may also remove aluminum complexes and remaining flocs, so in the future the team may decide to find a way to measure the level of aluminum in the effluent and employ this filtration step to see if the findings from previous studies can be matched.\n",
        "\n",
        "A redesign of the system is one of the team's main considerations going forward. While the system that uses flocculation and sedimentation removes some fluoride, many of the trials show that it often does not remove enough. The redesign will likely be first tested on the Automated system with pumps and later will be implemented into a gravity-powered system to be used in the field and for further research.\n",
        "\n",
        "In terms of making this process more accessible to a plant operator, the team will continue to explore and discuss ways in which successful treatment of the water can be visibly confirmed, since the fluoride ions as well as flocs are currently clear and difficult to see without a flashlight. The team hypothesizes that adding the clay to the treatment process may increase the turbidity of the flocs and thus give plant operators a concrete way of determining that the effluent water is safe to drink. This hypothesis will be tested within the next few weeks as well.\n",
        "\n",
        "Despite the progress that the AguaClara Fluoride team has made over the past three years, there still remain many unknown aspects of fluoride treatment that must be addressed before the process can be used in the field."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZXUik5SH-YFM",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Bibliography\"></a>\n",
        "## Bibliography\n",
        "(2013). Aluminum Fact Sheet. Retrieved from https://www.wqa.org/Portals/0/Technical/Technical Fact Sheets/2014_Aluminum.pdf\n",
        "\n",
        "Adsorption Equilibria. Retrieved from http://mimoza.marmara.edu.tr/~zehra.can/ENVE401/3. Adsorption Equilibria.pdf\n",
        "\n",
        "AguaClara Textbook. Retrieved from https://aguaclara.github.io/Textbook/index.html\n",
        "\n",
        "Akpan, P., Mehrabyan, T,. Sausele, D., & Zhang, V. (2018). Fluoride, Spring 2018. Retrieved from https://github.com/AguaClara/Fluoride-Auto/blob/master/FluorideReportSp18.md.\n",
        "\n",
        "Bhattacharya, P., & Samal, A. (2018, April). Fluoride contamination in groundwater, soil and cultivated foodstuffs of\n",
        "India and its associated health risks: A Review. Research Journal of Recent Sciences, 7(4),36-47. Retrieved from http://www.isca.in/rjrs/archive/v7/i4/6.ISCA-RJRS-2018-028.pdf.\n",
        "\n",
        "Brindha, K., Rajesh, R., Murugan, R., & Elango, L. (2010, February 20). Fluoride contamination in groundwater in parts of Nalgonda District, Andhra Pradesh, India. Retrieved from https://link.springer.com/article/10.1007/s10661-010-1348-0\n",
        "\n",
        "Cheng, M., Longo, A., & Vidal, B. (2016). Fluoride, Fall 2016. Retrieved from https://drive.google.com/file/d/0Bzej1vulT_XyWHZyLXdIWlk1RkE/view.\n",
        "\n",
        "Dubey, S., Agarwal, M., & Gupta, A. (2018). Experimental investigation of Al-F species formation and transformation during coagulation for fluoride removal using alum and PACl. Journal of Molecular Liquids, 266, 349–360. doi: 10.1016/j.molliq.2018.06.080\n",
        "\n",
        "EPA. “Water Treatability Database.” Environmental Protection Agency, 5 Feb. 2007. Retrieved from https://oaspub.epa.gov/tdb/pages/treatment/treatmentOverview.do?treatmentProcessId=1934681921\n",
        "\n",
        "Fawell, J., Bailey, K., Chilton, J., Dahi, E., Fewtrell, L., Magara, Y. Fluoride in Drinking-Water. IWA, 2006. Retrieved from https://www.who.int/water_sanitation_health/publications/fluoride_drinking_water_full.pdf\n",
        "\n",
        "Foo, K.Y. and Hameed, B.H. Insights into the modeling of adsorption isotherm systems, Chemical Engineering Journal, Volume 156, Issue 1, 2010, Pages 2-10, ISSN 1385-8947, https://doi.org/10.1016/j.cej.2009.09.013. Retrieved from https://www.sciencedirect.com/science/article/pii/S1385894709006147?via%3Dihub\n",
        "\n",
        "Garland, C., Weber-Shirk, M., Lion, L.W. Environmental Engineering Science. Feb 2016. ahead of print: http://doi.org/10.1089/ees.2015.0314\n",
        "Retrieved from https://www.liebertpub.com/doi/10.1089/ees.2015.0314\n",
        "\n",
        "Gebbie, P. (2001, December 5). Using Polyaluminum Coagulants in Water Treatment. Retrieved from http://wioa.org.au/conference_papers/2001/pdf/paper6.pdf\n",
        "\n",
        "Gong, W.-X., Qu, J.-hui, Liu, R.-P., & Lan, H.-C. (2011, December 16). Effect of aluminum fluoride complexation on fluoride removal by coagulation. Retrieved from https://www.sciencedirect.com/science/article/pii/S0927775711007564\n",
        "\n",
        "Grasso, D., Louie, M., Sausele, D., & Spiek, E. (2019). Fluoride Automated System, Spring 2019. Retrieved from https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202019/Fluoride-Auto-Spring-2019-Report.md.\n",
        "\n",
        "Hem, J. D., & Roberson, C. E. (1967). Form and Stability of Aluminum Hydroxide Complexes in Dilute Solution [PDF]. Washington: U.S. Department of the Interior.\n",
        "\n",
        "Ion-Selective Electrode Determination of Complex Formation Constants in Submicromolar Silver [PDF]. (1975). Analytical Chemistry.\n",
        "\n",
        "Huang, S., Jin, C., Kohli, S., Fluoride Gravity, Spring 2019. Retrieved from https://github.com/AguaClara/Fluoride_Gravity/blob/master/Spring%202019/Spring%202019%20Report.md.\n",
        "\n",
        "Jagtap, S., Yenkie M. K., Labhsetwar N., and Rayalu S. Chemical Reviews 2012 112 (4), 2454-2466\n",
        "DOI: 10.1021/cr2002855. Retrieved from https://pubs.acs.org/doi/pdf/10.1021/cr2002855\n",
        "\n",
        "LeChevallier, M. W., & Au, K. (2014, December 10). Water treatment and pathogen control: Process efficiency in achieving safe drinking water. Retrieved from https://www.who.int/water_sanitation_health/publications/9241562552/en/\n",
        "\n",
        "Lin, W. W.; Sung, S. S.; Chen, L. C.; Chung, H. Y.; Wang, C. C.; Wu, R. M.; Lee, D. J.; Huang, Chihpin; Juang, R. S.; Peng, X. F. and Chang, Hsi-Lih (2004). \"Treating high-turbidity water using full-scale floc blanket clarifiers\". Journal of Environmental Engineering, vol. 130, No. 12 (December), pp. 1481-1487. Retrieved from https://ascelibrary.org/doi/10.1061/%28ASCE%290733-9372%282004%29130%3A12%281481%29\n",
        "\n",
        "Logan, B. E., Hermanowicz, S. W., & Parker,A. S. (1987). A Fundamental Model for Trickling Filter Process Design. Journal (Water Pollution Control Federation), 59(12), 1029–1042.\n",
        "\n",
        "Longo, A., Li-Vidal, B., Cheng, M., & Zhang, V. (2017). Fluoride Floc Blanket, Spring 2017. Retrieved from https://github.com/AguaClara/Fluoride-Auto/blob/master/Spring%202017/Fluoride%20Spring%202017.pdf\n",
        "\n",
        "Longo, A., Zhang, V., & Cheng, M.\n",
        "\n",
        "Pang, C., Sarmiento, K., & Tsang, C. (2018). Fluoride Gravity, Fall 2018. Retrieved from https://github.com/AguaClara/Fluoride_Gravity/blob/master/Fall%202018/Fluoride_Grav_Fall2018_Report.md.\n",
        "\n",
        "Pang, C., Sarmiento, K., & Tsang, C. (2018). Fluoride Gravity, Summer 2018. Retrieved from https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202018/Fluoride_Summer2018.md.\n",
        "\n",
        "Rao, N. S., & Devadas, D. J. (2003, September 05). Fluoride incidence in groundwater in an area of Peninsular India. Retrieved from https://link.springer.com/article/10.1007/s00254-003-0873-3\n",
        "\n",
        "Savenko, A.V. Water Resources (2001) 28: 274. https://doi.org/10.1023/A:101049660675. Retrieved from https://link.springer.com/article/10.1023%2FA%3A1010496606751\n",
        "\n",
        "Su, S.T., Wu, R.M., Lee, D.J. Blanket dynamics in upflow suspended bed, Water Research, Volume 38, Issue 1, 2004, Pages 89-96, ISSN 0043-1354, https://doi.org/10.1016/j.watres.2003.09.002. Retrieved from http://www.sciencedirect.com/science/article/pii/S0043135403005025\n",
        "\n",
        "United States, NYC Environmental Protection. (2016). New York City 2016 Drinking Water Supply and Quality Report. New York. Retrieved from http://www.nyc.gov/html/dep/pdf/wsstate16.pdf.\n",
        "\n",
        "Weber-Shirk, Monroe, et al. AguaClara Textbook. AguaClara Cornell, 2018, aguaclara.github.io/Textbook/index.html.\n",
        "\n",
        "World Health Organization. (2016, August 29). Water-related diseases. Retrieved from http://www.who.int/water_sanitation_health/diseases-risks/diseases/fluorosis/en/.\n",
        "\n",
        "World Health Organization. Water, Sanitation and Health Team. (‎2004)‎. Guidelines for drinking-water quality. Vol. 1, Recommendations, 3rd ed. World Health Organization. Retrieved from https://apps.who.int/iris/handle/10665/42852\n",
        "\n",
        "Zan He, Huachun Lan, Wenxin Gong, Ruiping Liu, Yuping Gao, Huijuan Liu, Jiuhui Qu, Coagulation behaviors of aluminum salts towards fluoride: Significance of aluminum speciation and transformation, Separation and Purification Technology, Volume 165, 2016, Pages 137-144, ISSN 1383-5866, https://doi.org/10.1016/j.seppur.2016.01.017.\n",
        "Retrieved from https://www.sciencedirect.com/science/article/pii/S138358661630017X\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EvBA8kN4-dty",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Manual\"></a>\n",
        "# Manual"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vQiV3jwELreb",
        "colab_type": "text"
      },
      "source": [
        "###Fabricating the Sedimentation Tube\n",
        "*Safety Precautions: Wear heat-resistant gloves while using the PVC welder. Apply primer and cement glue in close proximity to the fume hood.*\n",
        "1.   Cut a 86.5 cm length of 1” inner diameter PVC, first using a hand saw and then using a band saw for accuracy. Lightly use the deburring tool to remove any sharp edges.\n",
        "2.   Measure the 50 cm point on the tube and indicate it with a marker. Also make two marks on either side of this point, in about 1-2 cm increments.\n",
        "<center>\n",
        " <img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/sed%20tube%20marks.JPG\" height=\"300\">\n",
        "  \n",
        "  **Figure 42:** Approximate areas to mark on the PVC before bending.\n",
        "    </center>\n",
        "  \n",
        "3.   Cap one end of the tubing and fill the entire tube with sand through the open end. Cap that end as well.\n",
        "4. Place the PVC horizontally in a vise. Line up the tube so that one of the marks on the end is touching the edge of the vise.\n",
        "5.   Use a protractor to measure a 160° angle, using the 50 cm point as the center point. Affix a clamp to the edge of the table to indicate the angle (where the angled part will reach once bent).\n",
        "\n",
        " <center>\n",
        "   <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Sed%20Tube%20Angle.png\" height =\"400\">\n",
        "  \n",
        "**Figure 43:** Method for measuring and marking an angle for bending the sedimentation tube.\n",
        "    </center>\n",
        "    \n",
        "6.   Make sure the top of the welding tool is open, so that air will blow out of the circular part as shown below. Allow the welding tool to heat up to around 400°F. \n",
        "\n",
        "<center>\n",
        " <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Welder.png\" height =\"250\">\n",
        "  \n",
        "**Figure 44:** Welder without the top screw-on piece.\n",
        "    </center>\n",
        "    \n",
        "7.   Heat up the PVC along the section indicated by the tick marks drawn earlier, making sure to heat all along the circumference of the pipe. \n",
        "8.   Slowly bend the PVC towards the clamp until it touches the clamp. \n",
        "9.   Hold the PVC in the bent position for several minutes, until it cools.\n",
        "10.   Cut a 40 cm length of 3/4\" PVC for the floc weir. \n",
        "11. Use the band saw and protractor to cut one end of the floc weir at an angle less than 70°. Any angle should work as long as the floc weir is not horizontal.\n",
        "12. The hole for the floc weir should be drilled on the 36.5 cm part of the 1\" PVC, 9.4 cm away from the bend. Angle the drill press so that it will drill at a 60-70° angle. \n",
        "13. Fit the 3/4\" PVC into the hole drilled in the 1\" PVC. Re-drilling may be necessary if the hole is too small. Also make sure that the floc weir does not protrude into the inner diameter of the 1\" PVC (filing may be necessary to prevent this).\n",
        "14.  Brush Clear Primer onto the angled end of the floc weir and around/inside the drilled hole of the sedimentation tube. \n",
        "15.  Apply Heavy Duty Clear PVC Cement to the angled end of the floc weir and the hole of the sedimentation tube. \n",
        "16.  Affix the floc weir to the sedimentation tube and allow it to sit in the right position for about two hours.\n",
        "17.  Run water through the tube and check for leaks. If there are leaks, use the PVC welder and clear welding rod to make the tube water tight.\n",
        "  \n",
        "*Note: The PVC welding rod should melt around 530°F. If the rod does not seem to be melting, use a screwdriver to clean out the screw-on top piece of the welder or take it to the Machine Shop to melt any remaining plastic using a blowtorch."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jlu4xbxu-Bdr",
        "colab_type": "text"
      },
      "source": [
        "### Calibrating the Fluoride Probe\n",
        "#### Set Up the Calibration Spreadsheet\n",
        "In Excel, set up a table as shown in the image below. The yellow highlighted cells are for the voltages measured during calibration, while the green highlighted cell is for a voltage measured during an experiment. A new spreadsheet should be used for every calibration. A linear graph, while not necessary, can be a helpful visual.\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Calibration%20Sheet.png\" height =\"300\">\n",
        "  \n",
        "**Figure 45:**  Calibration Excel Sheet.\n",
        "    </center>\n",
        "\n",
        "#### Using the Probe in ProCoDA\n",
        "The probe only needs to be added to ProCoDA once. On the \"Configuration\" tab, click the icon labeled \"Volts.\"\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Volts%20Link.png\" >\n",
        "  \n",
        "  **Figure 46:**  Volts icon on ProCoDA.\n",
        "    </center>\n",
        "  \n",
        "  Click \"Insert Sensor\" and name it \"Fluoride Probe.\" This screen is also a good way to check if the probe is reading a voltage.\n",
        "  \n",
        "  <center>\n",
        "    <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/Volt%20Reading.png?raw=true\">\n",
        "  \n",
        " **Figure 47:** Screen to add a probe and check voltage reading on ProCoDA.\n",
        "    </center>\n",
        "\n",
        "#### Daily Probe Calibration Procedure\n",
        "*Safety Precautions: Wear Latex gloves when handling fluoride, in order to prevent contact with bare skin.*\n",
        "1.  Make fluoride standard solutions of 1 mg/L, 5 mg/L, 10 mg/L, 20 mg/L, and 100 mg/L of fluoride. Use water from the tap (blue line). Standards do not need to be made daily.\n",
        "\n",
        "The following table is applicable for making 100 mL of standard solution from the 10,000 ppm F- standard solution.\n",
        "<table>\n",
        "  <tr>\n",
        "    <th> Standard Concentration (mg/L)</th>\n",
        "    <th>Volume of Fluoride Stock (microliters)</th>\n",
        "  <tr>\n",
        "    <td>1 </td>\n",
        "    <td>10</td>\n",
        "  </tr>\n",
        "  <tr>\n",
        "    <td>5</td>\n",
        "    <td> 50</td>\n",
        "  </tr>\n",
        "  <tr>\n",
        "    <td>10</td>\n",
        "    <td>100</td>\n",
        "  </tr>\n",
        "    <tr>\n",
        "    <td>20</td>\n",
        "    <td>200</td>\n",
        "  </tr>\n",
        "   <tr>\n",
        "    <td>100</td>\n",
        "    <td>1000</td>\n",
        "     \n",
        "  </tr>\n",
        "  \n",
        "</table>\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Standard%20Solutions.JPG\">\n",
        "  \n",
        "**Figure 48:**  The Summer 2019 standard solutions. The probe is stored in the 10 mg/L standard when not in use.\n",
        "    </center>\n",
        "    \n",
        "2.  Use a squeeze bottle to rinse the sides and bottom of the probe with DI water. Gently dab the probe dry with a Kimwipe.\n",
        "3.  Shake the standard (a good strategy is start with the 1 mg/L or 20 mg/L standard and go in order of increasing or decreasing concentration).\n",
        "4.  Push the O-ring down the probe until the fill hole is exposed.\n",
        "5.  Slide the probe into the hole in the bottle cap. Twist the cap onto the standard. Make sure the probe is not touching the sides or bottom of the bottle.\n",
        "6.  Click the \"Graphs\" tab on ProCoDA and highlight \"Fluoride Probe.\"\n",
        "7.  Do not touch the bottle or probe while the reading stabilizes. Try not to have any stirrers or pumps on at the same time, as the motion may disturb the probe. The calibration curve should look similar to the one below.\n",
        "\n",
        "<center>\n",
        " <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Probe%20Calibration%20Curve.png\">\n",
        "\n",
        "**Figure 49:** A typical calibration curve for a working probe.\n",
        "  \n",
        "  </center>\n",
        "  \n",
        "8.  Once the number under the box labeled \"Voltage\" (highlighted in Figure 31) stops changing, record the number next to the corresponding concentration in the Excel spreadsheet.\n",
        "9.  Repeat steps 2-8 for all standards.\n",
        "10. When the probe is not in use, store in the 10 mg/L standard. Remember to close the fill hole when finished with the probe for the day."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fky6QAi-LuZx",
        "colab_type": "text"
      },
      "source": [
        "###Troubleshooting the Fluoride Probe\n",
        "<table>\n",
        "  <tr>\n",
        "    <th> Issue</th>\n",
        "    <th>Type of Issue</th>\n",
        "  <tr>\n",
        "    <td>The probe does not respond to movement, such as shaking \n",
        "      <br>or stirring of the solution it is submerged in.</td>\n",
        "    <td>Likely a probe issue.</td>\n",
        "  </tr>\n",
        "  <tr>\n",
        "    <td> The probe reads a voltage of 0.600 mV.</td>\n",
        "    <td> Likely a computer or other hardware issue.</td>\n",
        "  </tr>\n",
        "  <tr>\n",
        "    <td>The reading does not change for standard solutions of different concentrations.</td>\n",
        "    <td>Probe or computer/hardware issue.</td>\n",
        "  </tr>\n",
        "    <tr>\n",
        "    <td>The R^2 value is below 0.99.</td>\n",
        "    <td>Likely a probe issue.</td>\n",
        "  </tr>\n",
        "   <tr>\n",
        "    <td> The probe fluctuates.</td>\n",
        "    <td>Likely a probe issue.</td>\n",
        "     \n",
        "  </tr>\n",
        "  \n",
        "</table>\n",
        "\n",
        "####Computer and Other Hardware Issues\n",
        "* Close out of ProCoDA and restart it. \n",
        "* Unplug the probe's BNC connector from the relay box (see Figure 32), wait ten seconds, replug. \n",
        "* Unplug the wire connecting the relay box to the ProCoDA box (see Figure 33), wait ten seconds, replug.\n",
        "* Replace the wire connecting the relay box to the ProCoDA box.\n",
        "\n",
        "####Probe Issues\n",
        "* Gently brush the sides of the probe with a soft toothbrush and toothpaste that contains fluoride. If needed, make a solution of some toothpaste and DI water and use a pipette to rinse the bottom of the probe (sensing area). Allow the probe to sit in DI water for five minutes, then in 1 mg/L fluoride solution for five minutes. If the graph still does not stabilize, put the probe in 100 mg/L fluoride solution until it stabilizes.\n",
        "* For a fluctuating probe, as long as the meter being used is a pH meter (the relay boxes in the lab that were manufactured in the machine shop should be voltmeters that work with pH probes), store the probe in buffer of 4.0 pH in lukewarm water (a little above room temperature). After about half an hour, stir the probe in the water and gently tap the side of the glass. Then whip the probe in one hand like with a thermometer.\n",
        "* Remake the standard solutions and try re-calibrating.\n",
        "* Stop using the probe and allow it to sit in 10 mg/L fluoride solution overnight. Check it the next day for better results.\n",
        "* Call the manufacturer for other troubleshooting procedures.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Relay%20Box.png\">\n",
        "  \n",
        "**Figure 50:**  Relay box.\n",
        "    </center>\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/ProCoDA%20Box.png\">\n",
        " \n",
        "**Figure 51:**  ProCoDA box.\n",
        "    </center>\n",
        "  \n",
        "  #### Manufacturer Troubleshooting Guide\n",
        "  <center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/Troubleshooting%202.png?raw=true\">\n",
        "  \n",
        "  **Figure 52:**  Manufacturer troubleshooting recommmendations.\n",
        "    </center>\n",
        "  \n",
        "  <center>\n",
        "  <img src=\"https://github.com/AguaClara/Fluoride-Auto/blob/master/Summer%202019/Images/Troubleshooting%201.png?raw=true\">\n",
        "    \n",
        "  **Figure 53:**  Manufacturer troubleshooting recommmendations.\n",
        "    </center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4qxbaI3-L1dp",
        "colab_type": "text"
      },
      "source": [
        "###Running Experiments in ProCoDA <a name=\"ProCoDA\"></a>\n",
        "1.   Before running any experiments, make sure the pump is calibrated on ProCoDA by clicking the \"Configuration\" tab and the icon under \"Pump Calibration\" on the right side. Pump calibration only needs to be done once.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Pump%20Calibration.png\">\n",
        "  \n",
        "**Figure 54:** Pump calibration icon on ProCoDA.\n",
        "    </center>\n",
        "    \n",
        "2.   Choose experimental concentrations for the reactor (concentration of PACl, fluoride, and clay desired in the reactor) and upflow velocity. Use the Python code (Code for Automated Experimental Flow Rates) to generate required flow rates in mL/s.\n",
        "3.  Type PACl flow rate into set point named \"Flow Rate of Coagulant.\"\n",
        "4.  Under \"Coagulant Pump\" set point, select \"Flow Rate of Coagulant\" and \"Coagulant mL/rev.\"\n",
        "5.  Type water flow rate into set point named \"Flow Rate of Water.\"\n",
        "6.  Under \"Water Pump\" set point, select \"Flow Rate of Water\" and \"Water mL/rev\" (or \"Water Tubing Size,\" depending on the required set points specified at the bottom of the screen).\n",
        "7.  Manually set the fluoride and clay pumps to the RPM generated by the Python code.\n",
        "8.  Under the \"Run Experiment\" rule, set Pump 1 and Pump 0 to the corresponding pumps (water and PACl). Make sure the pumps are set to \"ON.\" The direction of the pump should be \"ON\" or \"OFF\" depending on whether the pump should be running clockwise or counterclockwise.\n",
        "9.  Under the \"Process Operation\" tab, click the drop down arrow under \"Operator Selected State\" and click \"Run Experiment.\"\n",
        "10.  Turn on manual pumps.\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Run%20Experiment.png\">\n",
        "  \n",
        "**Figure 55:** ProCoDA menu for running experiment once all set points and rules are made.\n",
        "    </center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CXrgSYivAoEr",
        "colab_type": "text"
      },
      "source": [
        "## Experimental Checklist\n",
        "*Make sure this checklist is completed before running an experiment.*\n",
        "1.  All pumps are on. 100 RPM pumps that are controlled by ProCoDA should be set to \"Ext\" (External) while 600 RPM pumps that are controlled by ProCoDA should be set to \"mA.\"\n",
        "2.  All stock tanks have their stirrers on.\n",
        "3.  Turbidimeter is turned on.\n",
        "4.  Floc weir valve is closed.\n",
        "5.  Datalog is on. On the \"Configuration\" tab of ProCoDA, make sure there is a path in the white box labelled \"Datalog directory path.\" Alternatively, a text log (icon below the \"Datalog directory path\" icon that is labelled \"123 log\") can be used for short experiments, i.e. experiments that will not last more than two hours."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GvqVo0A4ykoJ",
        "colab_type": "text"
      },
      "source": [
        "### Emptying the Floc Weir\n",
        "*This procedure can be completed once the floc weir has been filled to capacity.*\n",
        "\n",
        "1. Remove the end tubing from the valve that is connected to the bottom of the weir.\n",
        "2. Fill a container partially with water from the sink (enough water to submerge the tubing).\n",
        "3. Submerge the tubing in the water and make sure it is completely filled. Without allowing any air to enter the tubing, re-attach it to the valve.\n",
        "4. Make sure the bottom/open end of the tubing is immersed in water. Open the floc weir, and as soon as water is seen flowing downwards, the end of the tubing can be taken out of the water. \n",
        "5. Use a flashlight to follow the level of the floc blanket as it drains.\n",
        "6. Close the valve, measure the concentration of the sample if desired, and dispose of the waste down the sink."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8xeWdOBu1fVz",
        "colab_type": "text"
      },
      "source": [
        "### Taking an Influent Sample\n",
        "*This procedure should be completed at the end of the experiment, after taking effluent and floc weir samples, since the team is unsure of its effects on the rest of the apparatus (flocculation and floc blanket).*\n",
        "1. Turn the PACl pump off but keep the PACl Pump off but keep the Water Pump and Fluoride Pump running.\n",
        "2. Open the blue valve over a waste container.\n",
        "3. Close the gray valve. \n",
        "4. Allow the flow to flow into the waste container until it reaches steady state, i.e. the flow becomes a steady stream of drips.\n",
        "5. Fill a new container with the influent sample.\n",
        "6. Once the influent collection is finished, open the gray valve.\n",
        "7. Close the blue valve.\n",
        "\n",
        "\n",
        "<center>\n",
        "  <img src = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Influent%20Valve.JPG\" height = \"300\">\n",
        "   \n",
        "**Figure 56:** Valve directly after the addition of fluoride in the system, used to collect samples of influent water.\n",
        "  </center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cv06b20oAj1G",
        "colab_type": "text"
      },
      "source": [
        "### Cleaning Procedure\n",
        "*This procedure should be completed after every experiment.*\n",
        "1. Pause all pumps.\n",
        "2. Manually set the Water Pump to the \"Int\" (Internal) setting. \n",
        "3. Increase the flow rate on the Water Pump to between 300 and 400 rpm. Turn the pump on.\n",
        "4. Allow the pump to run for several minutes. If the experiment contained fluoride, use a flashlight to ensure that flocs leave the top of the sedimentation tube.\n",
        "5. When the turbidity begins decreasing again, turn off the pump. \n",
        "6. Open the floc weir over a beaker or bottle and make sure all the flocs exit.\n",
        "7. If the reactor needs to be drained for any reason, place a container underneath it and disconnect the influent tubing at the bottom of the sedimentation tube. Then disconnect the effluent tubing at the top of the reactor. The floc weir can be opened and drained last."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BiTsuHFrGAm5",
        "colab_type": "text"
      },
      "source": [
        "## ProCoDA Method File"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CnTgFhHEGE_f",
        "colab_type": "text"
      },
      "source": [
        "### States\n",
        "- **OFF**: All sensors and pumps are turned off.\n",
        "- **JUST WATER**: Only the Just Water Pump is ON. Used for cleaning out the system after an experiment. Alternatively, the water pump can be set to \"Int\" (Internal) and the system can be manually cleaned out.\n",
        "- **RUN EXPERIMENT**: The Water Pump and Coagulant Pump are ON."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ks6ALZO2GH_E",
        "colab_type": "text"
      },
      "source": [
        "### Set Points\n",
        "Set points should be made in the order that they are used. For example, a variable that depends on a constant should appear after the constant in the list. Additionally, if any set points are made or deleted anywhere besides at the end of the list, check that the variables are set to depend on the correct constant since the selected set points depend on their position on the list. \n",
        "\n",
        "####Water\n",
        "Constants:\n",
        "* *Flow Rate of Water:* Flow rate of water in mL/s.\n",
        "* *Water mL/rev:* Based on tubing size. The Summer 2019 team used tubing size code 16, which corresponds to 0.8 mL/rev. Flow rates for other tubing sizes can be found under the Auto Tutorial for Peristaltic Pumps on the AguaClara Confluence page. Alternatively, the \"Water Tubing Size\" set point can be used. \n",
        "\n",
        "***Note regarding mL/rev:*** The Summer 2019 team volumetrically measured output from the peristaltic pumps and noticed that the actual mL/rev value was different than the one specified by the pump. The measured mL/rev value should be used for more accuracy, should there be any discrepancy in flow rates or solute concentrations.\n",
        "\n",
        "Variables:\n",
        "* *Water Pump:* Depends on the previous two constants. The number displayed in the \"Value\" box is the number that will be displayed on the pump divided by a factor of 100.\n",
        "\n",
        "<center>\n",
        "  <img src=\"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Images/Water%20Set%20Point.png\">\n",
        "  \n",
        "  **Figure 57:** Water Pump set point screen on ProCoDA.\n",
        "    </center>\n",
        "\n",
        "####PACl\n",
        "Constants:\n",
        "* *Flow Rate of Coagulant:* Flow rate of PACl in mL/s.\n",
        "* *Coagulant mL/rev:* Based on tubing size. The Summer 2019 team used orange-yellow microtubing, which corresponds to 0.019 mL/rev.\n",
        "\n",
        "Variables:\n",
        "* *Coagulant Pump:* Depends on the previous two constants.\n",
        "\n",
        "####Fluoride\n",
        "The team did not use the fluoride set point as part of the experiments since the pump was run manually. The following set points can be used to check the correct RPM at which the pump should be set, although it they are not necessary since there is available Python code for the same calculations.\n",
        "  Constants:\n",
        "* *Flow Rate of Fluoride:* Flow rate of fluoride in mL/s.\n",
        "* *Fluoride mL/rev:* Based on tubing size. The Summer 2019 team used orange-yellow microtubing.\n",
        "\n",
        "Variables:\n",
        "* *Fluoride Pump:* Depends on the previous two constants.\n",
        "\n",
        "####Clay\n",
        "The team did not use the clay set point as part of the experiments since the pump was run manually. The following set points can be used to check the correct RPM at which the pump should be set, although it they are not necessary since there is available Python code for the same calculations.\n",
        "  Constants:\n",
        "* *Flow Rate of Clay:* Flow rate of clay in mL/s.\n",
        "* *Water mL/rev:* Based on tubing size. The Summer 2019 team used orange-yellow microtubing.\n",
        "\n",
        "Variables:\n",
        "* *Clay Pump:* Depends on the previous two constants.\n",
        "\n",
        "####Turbidity\n",
        "Make sure this set point is correctly implemented so that turbidity will be logged in the datalog.\n",
        "\n",
        " Constants:\n",
        "  * *Com Port Number:* Should be set to 1.\n",
        "  * *Turbidity ID:* Should be set to 1.\n",
        "  \n",
        "Variables:\n",
        "  * *Turbidity:* Depends on the previous two constants."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LaELphDxGIE-",
        "colab_type": "text"
      },
      "source": [
        "<a name=\"Python_Code\"></a>\n",
        "## Python Code"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uNiQXErgGRg7",
        "colab_type": "text"
      },
      "source": [
        "### Code for Automated Experimental Flow Rates\n",
        "The following code is used to calculate experimental flow rates for desired concentrations of PACl, fluoride, and clay. The PACl flow rate in mL/s can be put into ProCoDA, while the fluoride and clay flow rates are most useful in RPM, since these pumps are run manually."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cq2qbVl1GpO3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip install aguaclara"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "anc_Yv0I8Goq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import math as m\n",
        "import numpy as np\n",
        "from aguaclara.play import*\n",
        "\n",
        "#UPFLOW VELOCITY\n",
        "#Diameter of reactor tube\n",
        "D_reactor = 1 * u.inch\n",
        "#Input upflow velocity\n",
        "V_up = 1.5 * u.mm/u.s\n",
        "#Flow rate into reactor\n",
        "Q_reactor = (m.pi*((D_reactor/2)**2)*V_up).to(u.mL/u.s)\n",
        "#Height of reactor\n",
        "H_reactor = 50 * u.inch\n",
        "# Volume of reactor\n",
        "Vol_reactor = (m.pi * ((D_reactor)**2)*H_reactor/4).to(u.mL)\n",
        "# Residence time in reactor\n",
        "T_residence= (Vol_reactor/Q_reactor).to(u.minute)\n",
        "\n",
        "#PACL\n",
        "#Concentration of PACl stock\n",
        "C_stock_PACl = 1000*u.mg/u.L\n",
        "#Input concentration of PACl that you want in the reactor\n",
        "C_reactor_PAC = 25*u.mg/u.L\n",
        "#Coagulant microtubing (orange-yellow)\n",
        "oy_tube=0.019*u.mL/u.revolutions\n",
        "#Flow rate of PACl in mL/s (put into ProCoDA)\n",
        "Q_stock_PAC = Q_reactor*(C_reactor_PAC/C_stock_PACl)\n",
        "print('PACl should be set in ProCoDA to: '+ str(Q_stock_PAC))\n",
        "Q_stock_PAC_rpm = (Q_stock_PAC/oy_tube)*60*u.sec/u.min\n",
        "print('PACl in rpm is: '+str(Q_stock_PAC_rpm))\n",
        "\n",
        "#FLUORIDE\n",
        "#Concentration of fluoride stock\n",
        "C_stock_F = 1000*u.mg/u.L\n",
        "#Input concentration of fluoride that you want in the reactor\n",
        "C_reactor_F = 5*u.mg/u.L\n",
        "#Fluoride microtubing (orange-yellow)\n",
        "oy_tube=0.019*u.mL/u.revolutions\n",
        "#Flow rate of fluoride in mL/s (check in ProCoDA)\n",
        "Q_stock_F = Q_reactor*(C_reactor_F/C_stock_F)\n",
        "print('Fluoride can be set in ProCoDA to: '+str(Q_stock_F))\n",
        "#Flow rate of fluoride in RPM (put into manual pump)\n",
        "Q_stock_F_rpm = (Q_stock_F/oy_tube)*60*u.sec/u.min\n",
        "print('Fluoride pump should be set at: '+str(Q_stock_F_rpm))\n",
        "\n",
        "#CLAY\n",
        "#Concentration of clay stock\n",
        "C_stock_clay = 100*u.mg/u.L\n",
        "#Input concentration of PACl that you want in the reactor\n",
        "C_reactor_clay = 0*u.mg/u.L\n",
        "#Coagulant microtubing (orange-yellow)\n",
        "oy_tube=0.019*u.mL/u.revolutions\n",
        "#Flow rate of PACl in mL/s (put into ProCoDA)\n",
        "Q_stock_clay = Q_reactor*(C_reactor_clay/C_stock_clay)\n",
        "print('Clay should be set in ProCoDA to: '+ str(Q_stock_clay))\n",
        "Q_stock_clay_rpm = (Q_stock_clay/oy_tube)*60*u.sec/u.min\n",
        "print('Clay in rpm is: '+str(Q_stock_clay_rpm))\n",
        "\n",
        "#WATER\n",
        "#Flow rate of water in mL/s (put into ProCoDA)\n",
        "Q_tap = Q_reactor - 1*Q_stock_PAC - 1*Q_stock_F - 1*Q_stock_clay\n",
        "print('Water should be set in ProCoDA to: '+str(Q_tap))\n",
        "water_tube = 0.8*u.mL/u.revolutions\n",
        "Q_tap_rpm = (Q_tap/(water_tube))*(60*u.sec/u.minute)\n",
        "print('Water in rpm should be: '+str(Q_tap_rpm))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HUiAGJouPYTo",
        "colab_type": "text"
      },
      "source": [
        "### Code for Gravity Experimental Flow Rates <a name=\"Grav\"></a>\n",
        "The following code is used to calculate flow rates for upflow velocity and coagulant in the gravity-powered system.\n",
        "\n",
        "The slope is taken from data collected from the Spring 2019 Gravity team, which volumetrically measured the flow rate of PACl and the upflow velocity of the system and generated a model for calculating flow rates based on heights using these experimental values. The graph may need to be redone in the future because of previous inconsistencies in height measurements (some teams measured from the bottom of the adjustable maneuvering trays while others measured from the top)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tMOBhSSGQEAW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import math as m\n",
        "import numpy as np\n",
        "from aguaclara.play import*\n",
        "\n",
        "#UPFLOW VELOCITY\n",
        "#Cross-sectional area of\n",
        "A_reactor = 506.7*u.mm**2\n",
        "#Input height of effluent tube\n",
        "H_effluent = 2.5 * u.cm\n",
        "#Volumetric flow rate\n",
        "#Slope of the graph from Spring 2019\n",
        "Slope_effluent = 0.028778 * u.mL/u.s/u.cm\n",
        "Q_reactor = H_effluent * Slope_effluent\n",
        "Q_reactor = Q_reactor.to(u.mm**3/u.s)\n",
        "#Upflow Velocity\n",
        "V_up = Q_reactor/A_reactor\n",
        "print(\"Upflow velocity is \"+str(V_up))\n",
        "\n",
        "#PACL CONCENTRATION\n",
        "#Concentration of PACl stock\n",
        "C_stock_PACl = 1000 * u.mg/u.L\n",
        "#Input height of PACl adjustable maneuver\n",
        "H_PACl = 6 * u.cm\n",
        "#Volumetric flow rate\n",
        "#Slope of the graph from Spring 2019\n",
        "Slope_PACl = 0.000544 * u.mL/u.s/u.cm\n",
        "Q_stock_PAC = H_PACl * Slope_PACl\n",
        "C_stock_PAC = 1000 * u.mg/u.L\n",
        "C_stock_PAC = C_stock_PAC.to(u.mg/u.mL)\n",
        "C_reactor_PAC = Q_stock_PAC*C_stock_PAC/(Q_reactor.to(u.L/u.s))\n",
        "print(\"Concentration of PACl in the reactor is \"+str(C_reactor_PAC))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KIjxhJ1U_D7v",
        "colab_type": "text"
      },
      "source": [
        "### Code for Graphing Cumulative Data with the Langmuir Isotherm\n",
        "The following code was developed by the Spring 2019 team to graph all data since Summer 2018, along with a best fit line for the data and the theoretical curve that follows the Langmuir Isotherm."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w6D_Vlj_B46Y",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import scipy.stats as stats\n",
        "import numpy as np\n",
        "import scipy.optimize as opt\n",
        "\n",
        "# import the summer 2018 data: summer18\n",
        "summer18 = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Langmuir%20Isotherm%20Data/Summer%202018%20data%20points.csv')\n",
        "summer18_effluent=summer18.iloc[:,0]\n",
        "summer18_uptake=summer18.iloc[:,1]\n",
        "summer18plot, =plt.plot(summer18_effluent,summer18_uptake,\"ro\")\n",
        "\n",
        "# import the fall 2018 data: fall18\n",
        "fall18 = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Langmuir%20Isotherm%20Data/Fall%202018%20data.csv')\n",
        "fall18_effluent=fall18.iloc[:,0]\n",
        "fall18_uptake=fall18.iloc[:,1]\n",
        "fall18plot, =plt.plot(fall18_effluent,fall18_uptake,\"bo\")\n",
        "\n",
        "# import the spring 2019 data: spring19\n",
        "spring19 = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Langmuir%20Isotherm%20Data/Spring%202019%20data.csv')\n",
        "spring19_effluent=spring19.iloc[:,0]\n",
        "spring19_uptake=spring19.iloc[:,1]\n",
        "spring19plot, =plt.plot(spring19_effluent,spring19_uptake,\"ys\")\n",
        "\n",
        "# import the summer 2019 data: summer19\n",
        "summer19 = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Summer%202019%20data.csv')\n",
        "summer19_effluent=summer19.iloc[14,1]\n",
        "summer19_uptake=summer19.iloc[14,4]\n",
        "summer19plot, =plt.plot(summer19_effluent,summer19_uptake,\"ms\")\n",
        "\n",
        "# import the summer 2019 data without clay: summer19noclay\n",
        "summer19noclay = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/No%20Clay.csv')\n",
        "summer19_noclay_effluent=summer19noclay.iloc[:,1]\n",
        "summer19_noclay_uptake=summer19noclay.iloc[:,4]\n",
        "summer19noclayplot, =plt.plot(summer19_noclay_effluent,summer19_noclay_uptake,\"m*\")\n",
        "\n",
        "# import the summer 2019 data with clay: summer19clay\n",
        "summer19clay = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Clay.csv')\n",
        "summer19_clay_effluent=summer19clay.iloc[:,1]\n",
        "summer19_clay_uptake=summer19clay.iloc[:,4]\n",
        "summer19clayplot, =plt.plot(summer19_clay_effluent,summer19_clay_uptake,\"b*\")\n",
        "\n",
        "# find the logarithmic line of best fit for the data\n",
        "alldata=pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Langmuir%20Isotherm%20Data/All%20Experimental%20Data%20Points.csv')\n",
        "alldata_effluent=alldata.iloc[:,0]\n",
        "alldata_uptake=alldata.iloc[:,1]\n",
        "alldata_theoretical_uptake=alldata.iloc[:,2]\n",
        "np.polyfit(np.log(alldata_effluent),alldata_uptake,1)\n",
        "# line of best fit for data: uptake = 95.63250099 log(effluent) + 131.89094095\n",
        "\n",
        "# graph line of best fit with data points\n",
        "bestfit_effluent = np.array(np.linspace(0.45,12))\n",
        "bestfit_uptake =  95.63250099 * np.log(bestfit_effluent) + 131.89094095\n",
        "bestfit_plot, =plt.plot(bestfit_effluent,bestfit_uptake)\n",
        "\n",
        "# find r^2 value for line of best fit\n",
        "log_alldata_effluent=np.log(np.array(alldata_effluent))\n",
        "linreg = stats.linregress(log_alldata_effluent, alldata_uptake)\n",
        "slope, intercept, r_value = linreg[0:3]\n",
        "print(\"R-squared for line of best fit\", (r_value ** 2))\n",
        "\n",
        "# get theoretical Langmuir Isotherm\n",
        "theoretical = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Spring%202019/Langmuir%20Isotherm%20Data/theoretical%20data%20points.csv')\n",
        "theoretical_effluent=theoretical.iloc[:,0]\n",
        "theoretical_uptake=theoretical.iloc[:,1]\n",
        "theoreticalplot, =plt.plot(theoretical_effluent,theoretical_uptake,\"m\")\n",
        "\n",
        "# plot the data set\n",
        "plt.xlim(0, 11)\n",
        "plt.ylim(0,400)\n",
        "plt.xlabel(\"Effluent Fluoride Concentration (mg/L)\")\n",
        "plt.ylabel(\"Uptake (mg Fluoride/g PaCl)\")\n",
        "plt.legend((summer18plot, fall18plot, spring19plot, summer19noclayplot, summer19clayplot, bestfit_plot,theoreticalplot),(\"Summer 2018\", \"Fall 2018\", \"Spring 2019\", \"Summer 2019 No Clay\", \"Summer 2019 Clay\", \"Line of Best Fit\",\"Theoretical Langmuir Isotherm\"))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fu00n2wL-xWr",
        "colab_type": "text"
      },
      "source": [
        "### Code for Graphing Clay vs. No Clay (Summer 2019)\n",
        "The following code is used to graph the Summer 2019 data points, which are distinguished by whether clay was used in each experiment or not. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4MZmylWS-vDF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import scipy.stats as stats\n",
        "import numpy as np\n",
        "import scipy.optimize as opt\n",
        "\n",
        "# import the summer 2019 data without clay: summer19noclay\n",
        "summer19noclay = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/No%20Clay.csv')\n",
        "summer19_noclay_effluent=summer19noclay.iloc[:,1]\n",
        "summer19_noclay_uptake=summer19noclay.iloc[:,4]\n",
        "summer19noclayplot, =plt.plot(summer19_noclay_effluent,summer19_noclay_uptake,\"mo\")\n",
        "\n",
        "# import the summer 2019 data with clay: summer19clay\n",
        "summer19clay = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Clay.csv')\n",
        "summer19_clay_effluent=summer19clay.iloc[:,1]\n",
        "summer19_clay_uptake=summer19clay.iloc[:,4]\n",
        "summer19clayplot, =plt.plot(summer19_clay_effluent,summer19_clay_uptake,\"bo\")\n",
        "\n",
        "# plot the data set\n",
        "plt.title(\"Summer 2019 Uptake vs. Effluent Points\")\n",
        "plt.xlim(0,6)\n",
        "plt.ylim(0,200)\n",
        "plt.xlabel(\"Effluent Fluoride Concentration (mg/L)\")\n",
        "plt.ylabel(\"Uptake (mg Fluoride/g PaCl)\")\n",
        "plt.legend((summer19noclayplot, summer19clayplot),(\"No Clay\", \"Clay\"))\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WlIo1_DHCIlP",
        "colab_type": "text"
      },
      "source": [
        "### Code for Graphing Automated vs. Gravity (Summer 2019)\n",
        "The following code is used to graph the Summer 2019 data points, which are the successful experiments distinguished by the system that they were run in."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "iHRMSJLgCnKb",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import scipy.stats as stats\n",
        "import numpy as np\n",
        "import scipy.optimize as opt\n",
        "\n",
        "# import auto data\n",
        "summer19auto = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/Working%20Experiments%20Auto.csv')\n",
        "summer19_auto_effluent=summer19auto.iloc[:,1]\n",
        "summer19_auto_uptake=summer19auto.iloc[:,4]\n",
        "summer19autoplot, =plt.plot(summer19_auto_effluent,summer19_auto_uptake,\"rs\")\n",
        "\n",
        "# import grav data\n",
        "summer19grav = pd.read_csv('https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/Working%20Experiments%20Grav.csv')\n",
        "summer19_grav_effluent=summer19grav.iloc[:,1]\n",
        "summer19_grav_uptake=summer19grav.iloc[:,4]\n",
        "summer19gravplot, =plt.plot(summer19_grav_effluent,summer19_grav_uptake,\"cs\")\n",
        "\n",
        "plt.title(\"Summer 2019 Uptake vs. Effluent Successful Experiments\")\n",
        "plt.xlim(0,6)\n",
        "plt.ylim(0,175)\n",
        "plt.xlabel(\"Effluent Fluoride Concentration (mg/L)\")\n",
        "plt.ylabel(\"Uptake (mg Fluoride/g PaCl)\")\n",
        "plt.legend((summer19autoplot, summer19gravplot), (\"Automated System\", \"Gravity System\"))\n",
        "plt.savefig(\"Working Graphs\")\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KSpJ9bUMqDBM",
        "colab_type": "text"
      },
      "source": [
        "### Code for Graphing Fluoride Concentration and Turbidity over Time from a Datalog\n",
        "The following code is used to graph fluoride concentration and turbidity along the same x-axis, using a datalog from ProCoDA. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_tOrg1_tphrf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "```python\n",
        "import aguaclara.research.procoda_parser as pp\n",
        "import scipy.stats as stats\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import scipy as sp\n",
        "import scipy.optimize as opt\n",
        "from aguaclara.play import *\n",
        "\n",
        "#Graph fluoride concentration and turbidity vs. time\n",
        "#Import the datalog\n",
        "july10exp = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-10-19/datalog%207-10-2019.xls\"\n",
        "data_raw = pd.read_csv(july10exp, delimiter=\"\\t\")\n",
        "data = pp.remove_notes(data_raw)\n",
        "time = data.iloc[:,1]/60\n",
        "time_number = pd.to_numeric(time)\n",
        "fluoride = data.iloc[:,3]\n",
        "turbidity = data.iloc[:,5]\n",
        "\n",
        "# ax1 is the axis handle for the first y-axis\n",
        "fig, ax1 = plt.subplots()\n",
        "ax1.set_xlabel(\"Time (minutes)\")\n",
        "ax1.set_ylabel(\"Effluent Fluoride Concentration (mg/L)\")\n",
        "ax1.set_ylim(0,10)\n",
        "# line1 is the line handle for the effluent_turbidity graph\n",
        "fluoride_plot, = ax1.plot(time, fluoride, color=\"blue\")\n",
        "\n",
        "# ax2 is an axis handle for the second y-axis, with the same x-axis as ax1\n",
        "ax2 = ax1.twinx()\n",
        "ax2.set_ylabel(\"Effluent Turbidity (NTU)\")\n",
        "ax2.set_ylim(0,5)\n",
        "# turbidity_plot is the line handle for the effluent_turbidity graph\n",
        "turbidity_plot, = ax2.plot(time, turbidity, color=\"green\")\n",
        "\n",
        "plt.title(\"7-10-19 Effluent Fluoride Concentration and Turbidity\")\n",
        "plt.legend((fluoride_plot, turbidity_plot),(\"Fluoride Concentration\",\"Turbidity\"))\n",
        "plt.savefig(\"7-10-19 Effluent Fluoride Concentration and Turbidity\")\n",
        "\n",
        "```\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oNGfspqtsu3R",
        "colab_type": "text"
      },
      "source": [
        "### Code for Graphing Fluoride Concentration and Floc Blanket Height over Time from a Datalog\n",
        "The following code is used to graph fluoride concentration and height of the floc blanket along the same x-axis, using a datalog from ProCoDA. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rgcDzagZsJcU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "```python\n",
        "import aguaclara.research.procoda_parser as pp\n",
        "import scipy.stats as stats\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import scipy as sp\n",
        "import scipy.optimize as opt\n",
        "from aguaclara.play import *\n",
        "\n",
        "#Import the datalog\n",
        "july10exp = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-10-19/datalog%207-10-2019.xls\"\n",
        "data_raw = pd.read_csv(july10exp, delimiter=\"\\t\")\n",
        "data = pp.remove_notes(data_raw)\n",
        "time = data.iloc[:,1]/60\n",
        "time_number = pd.to_numeric(time)\n",
        "fluoride = data.iloc[:,3]\n",
        "floc_blanket = data.iloc[:,6]\n",
        "\n",
        "#Graph fluoride concentration and floc blanket height vs. time\n",
        "# ax3 is the axis handle for the effluent fluoride concentration\n",
        "fig, ax3 = plt.subplots()\n",
        "ax3.set_xlabel(\"Time (minutes)\")\n",
        "ax3.set_ylabel(\"Effluent Fluoride Concentration (mg/L)\")\n",
        "ax3.set_ylim(0,4)\n",
        "# fluoride_plot is the line handle for the effluent fluoride concentration graph\n",
        "fluoride_plot, = ax3.plot(time, fluoride, color=\"blue\")\n",
        "\n",
        "# ax4 is an axis handle for the second y-axis, with the same x-axis as ax3\n",
        "ax4 = ax3.twinx()\n",
        "ax4.set_ylabel(\"Floc Blanket Height (cm)\")\n",
        "ax4.set_ylim(0,90)\n",
        "# flocblanket_plot is the line handle for the height of the floc blanket\n",
        "flocblanket_plot, = ax4.plot(time, floc_blanket, \"ro\")\n",
        "\n",
        "plt.title(\"7-10-19 Floc Blanket Height and Fluoride Concentration over Time\")\n",
        "plt.legend((fluoride_plot, flocblanket_plot),(\"Fluoride Concentration (mg/L)\",\"Floc Blanket Height (cm)\"))\n",
        "plt.savefig(\"7-10-19 Floc Blanket Height and Fluoride Concentration\")\n",
        "\n",
        "```"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GkMVqqPvs2_B",
        "colab_type": "text"
      },
      "source": [
        "### Code for Graphing Turbidity and Floc Blanket Height over Time from a Datalog\n",
        "The following code is used to graph effluent turbidity and height of the floc blanket along the same x-axis, using a datalog from ProCoDA. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CjMJO3ERsRMj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "```python\n",
        "import aguaclara.research.procoda_parser as pp\n",
        "import scipy.stats as stats\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import scipy as sp\n",
        "import scipy.optimize as opt\n",
        "from aguaclara.play import *\n",
        "\n",
        "#Import the datalog\n",
        "july10exp = \"https://raw.githubusercontent.com/AguaClara/Fluoride-Auto/master/Summer%202019/Data%20and%20Graphs/7-10-19/datalog%207-10-2019.xls\"\n",
        "data_raw = pd.read_csv(july10exp, delimiter=\"\\t\")\n",
        "data = pp.remove_notes(data_raw)\n",
        "time = data.iloc[:,1]/60\n",
        "time_number = pd.to_numeric(time)\n",
        "fluoride = data.iloc[:,3]\n",
        "turbidity = data.iloc[:,5]\n",
        "floc_blanket = data.iloc[:,6]\n",
        "\n",
        "#Graph turbidity and floc blanket height vs. time\n",
        "fig, ax5 = plt.subplots()\n",
        "ax5.set_xlabel(\"Time (minutes)\")\n",
        "ax5.set_ylabel(\"Effluent Turbidity (NTU)\")\n",
        "ax5.set_ylim(0,4)\n",
        "# turbidity_plot is the line handle for the effluent turbidity graph\n",
        "turbidity_plot, = ax5.plot(time, turbidity, color=\"green\")\n",
        "\n",
        "ax6 = ax5.twinx()\n",
        "ax6.set_ylabel(\"Floc Blanket Height (cm)\")\n",
        "ax6.set_ylim(0,90)\n",
        "# flocblanket_plot is the line handle for the effluent_turbidity graph\n",
        "flocblanket_plot, = ax6.plot(time, floc_blanket, \"ro\")\n",
        "\n",
        "plt.title(\"7-10-19 Floc Blanket Height and Turbidity over Time\")\n",
        "plt.legend((fluoride_plot, flocblanket_plot),(\"Turbidity (NTU)\",\"Floc Blanket Height (cm)\"))\n",
        "plt.savefig(\"7-10-19 Turbidity and Floc Blanket Height\")\n",
        "\n",
        "```"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}