Improving Water Treatment Plant Performance With WTC-Coag

To develop a Water Treatment Plant Upgrade Business Case – A water treatment plant operator has heard of WTC-Coag and he would like to develop a business case to seek funding to purchase this software and implement it in his treatment plant to improve his treatment plant performance. He has a trial copy of WTC-Coag and he would like to use historical water quality data to assess the suitability of the software for his plant.

Treatment Plant description: Coagulation / flocculation - Dissolved air flotation – Granular media filtration – Chlorination. Chemical used: Aluminium sulphate / Sodium hydroxide / Chlorine gas / Powdered activated carbon / Fluorosilicic acid / Potassium permanganate

For this assignment, we don’t have a specific format of your submission, as far as you provide the answers of the questions below

Questions

Inspect treatment plant record using the provided Excel spreadsheet, WTP records.xlsx

1. Develop data pre-treatment procedure / data validation
2. Identify key parameters need for this modelling assignment, briefly state why your selected parameters are useful for this work, concentrate on those input parameters for WTC-Coag and other parameters required to perform various calculations and evaluations.Identify data errors in the Excel spreadsheet, particularly those easily observable errors, such as, out of range or data entry error. Provide comment on why you think they are incorrect.

1. Apply data correction methods, such as, how to handle missing data (average the two neighbourhood data points or fill the same value as the previous data point or exclude the missing data for calculation).
2. State clearly of how you setup / correct the data and the assumption of why you believe the way you handle it is acceptable, provide a brief justification of the assumption and impact of the result, write them as a section (few paragraphs) in your document. Each correct item (item with a good reason) will score 1 mark but each incorrect item will deduct 1 mark. You also need to label and comment what you have done in the Excel file, you may use a separate cell or the comment function in Excel to illustrate your workout in the sheet. Submit Excel file for assessment.

3. Assess plant operations data

1. Determine annual plant output in ML (megalitres) Financial Year (July to June)

1. Determine annual alum consumption in tons

Financial Year (July to June)

Determine predicted alum dose using the provided water quality data as inputs for WTC-Coag

5. Construct a graph similar to Fig 1 and provide interpretation of the suitability of the software for this treatment plant

6. Determine annual alum cost and potential saving after using WTC-Coag

You will be required to determine both current cost and potential cost saving after using WTC-Coag for the plant, you may provide the cost saving in a range including justifications of your assumption and discuss the limitations of your assessment.

7. Recommendation including comparison of other similar products

Provide recommendation of whether this product (WTC-Coag) is suitable for the plant, compare other similar products (web search) and discuss the suitability, including model inputs and functionalities.

This section should contain, a brief background, problem statement (such as, the product is intended to address under dosing of alum and over dosing of alum, expand along these subheadings), options considered, risk assessment and final recommendation.

The treatment plant records are for the period between 2010 and 2017. Although the data is available for all years between the specified period; 2010 and 2017 only have data for 3 months and 5 months respectively. The data is presented in different categories depending on the quantity being assessed. For instance, in order to obtain highly quality drinking water the unclean water is subjected through as series of purification steps that are classified under several categories such as, raw water, coagulated water, floated water, filtered water, transfer water, and treated water. The columns of data do not have the same number of values with some being highly populated and others having considerably few values in them e.g. Manganese columns have very few values in them. The assessment of some parameters is differentiated into laboratory and online; denoting the source that was used to get the values presented in the rows.   

Part a

From the data presented in the excel document it is clear there are some data values that are more important than others. As such, the necessary parameters to be used in the WTC-Coag assessment are as follows: Colour (COLOUR), Turbidity (TURB), Target coagulation pH, Enhance Coagulation (EC), UV absorbance at 254nm (UV),  Aluminium Sulphate (ALUM). The parameter aforementioned at the most important with regard to the computations relating to water purification The three parameters that will be employed in the calculation of WTC-Coag are UV, COLOUR, and TURB. Other parameters will be used in conjunction with the ones mentioned above to compute alkalinity, alum dose rate, % removal of coagulable DOC, and pH changes.

Part b

There are several data errors but the most notable ones relate to matters of data entry where none numeric values are entered e.g. "*" and "plant full". The use of non-numeric values could also be a misplacement error where a statement like "plant full" is supposed to be in the comment column as a justification of why not records are present for a given date. Overestimation error results in the entry of considerably large data values that are not consistent with other data values in the same category or column. Another error is the failure to use consistent units of measure: for instance, water is supposed to be assessed in mega litres there are however some instances where it is represented in litres, even thou it can be easily transcribed into mega litres. Lastly, there are numerous missing values; in some categories the missing values are considerably large, but in others they are few.

Part a

Part c

In the cases that none numeric values have been entered they should all be replaced by a value that is an average of the four neighbouring values in the same column or category (preferably the values should be from the same year). For large missing sets of data, specific column should be ignored completely. For individual missing values the specific row of that column can be ignored to avoid overestimation or underestimation of the actual value.

Part d

Irrelevant data: I will delete all rows with data for the period between 2010 and May 20 of 2014; so that we are left with three years of data for the period between May 21of 2014 and May 21 of 2017. This is appropriate to do because the assessment of the data needs to be done with regard to most recent three years of data.

Missing Data: I will delete columns with too many missing values because they will be inconsequential in the overall calculations. Moreover, some of the columns with missing values are unnecessary in the computation of key assessment parameters that will be employed in decision-making by WTC managers. In situations where the missing values are few and supposed to be numeric we will get an average for the empty cell using an average of four values; two values above the empty cell and two below it.  This will allow us to get a more realistic value the single missing values. In a situation where it is an entire row that is missing data the row will be completely ignored since the data for that data was not collected or transcribed into the treatment plan records.

Non-numerical data: all non numeric data that appears amongst numeric values will be replaced with a value that is the average of four points (2 above and 2 below the non-numeric value). By so doing the value the figure entered in place of the nonnumeric data will be considerably similar to the rest of the data.

Considerably large data (outliners): This will also be replaced by values that are averages of the neighbouring cell values. Outliners are important to eliminate because they negative affect the spread and centre values of the data set. In addition, outliners present an unrealistic representation of data distribution and the values that can be assumed by a variable. For instance, it is unrealistic for temperature to take up values of 30008?C; this could be a misrepresentation of 38?C.

Part b

Part a

The value for 3rd of February 2016 was considerably large indicating an error. An average figure of four neighbouring values was used to replace the error term (average of 2 values above plus 2 values below the error cell)

Part b

The aluminium consumption is calculated by multiplying the dose rate (plus drop test) by the volume of water in both tanks. The fraction errors with dose rate and drop test were corrected by taking numerator value to be for the dose rate and the denominator to be for the drop test e,g. 40/35 represents 40 dose rate and 35 drop test. A single underestimation error was observed for 0.45 mg which was adjusted to 45mg

The 10 points were selected randomly from the data with regard to the points which indicates the least absolute difference between Predicted Aluminium dose and Actual Aluminium dose

Judging from the data obtained from the website versus actual data retrieved from treatment plant records it is clear that there is considerable variation between the two sets of data as such it is easy to conclude that the utility of the website in a business establishment for decision-making purposes is very limited due to inaccuracy associated with underestimation and overestimation of values. Additional functions that can be employed would have to be assessment of measures of central tendency and dispersion for the two sets of data to establish without any doubt on the differences between actual data collected from plant operations and predicted values generate via the website platform. A hypothesis analysis can be performed whose null hypothesis is founded on the principle that there is no significant difference between the mean for predicted Aluminium dose and actual aluminium dose.

I only employed 4 percentages (100% to 70%) because I wanted to also include both predicted enhanced aluminium and actual aluminium in the chart without it being too crowded.  From the chart presented above one is able to clearly see the steps and movements of the individual lines for the time period between 2014 and 2017.

By comparing the costs associated with other water treatment procedures and the relatively cheap cost of Aluminium it is clear that the adoption of WTC-coag will greatly improve saving by mitigating costs. This reduction in costs is done through the substitution of expensive filtration and purification systems for a more cost effective technique in the utilization of aluminium. The general assumption is that we will not take into consideration the salvage value of previously used water treatment equipment, or the cost of machinery that will be employed in the aluminium dosing procedure. In addition, the variation of aluminium cost in the global market will be limited within acceptable parameters; as such, the price of aluminium will not be expected to be high than $5 per kg or less than $0.50 per kg.

Background

There are similar products in the market but WTC-coag is considered the most effective technique compared to its predecessors and market equivalents. It allows the establishment to use considerable less Aluminium in the dosing process but yields better treatment results.

Problem Statement

The main issue it the betterment of the treatment process without increasing the inputs used in the aluminium dosing processes of removal of DOC.

My recommendation will have to be employment of the WTC-coag system because it will greately improve the quality of the treated water by removing a higher percentage of DOC compared to tradition filtration and purification systems. Moreover, WTP