Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Robotics Radio. Generated commentary on the latest robotics and control papers.
Rosa: I'm Rosa.
Dev: And I'm Dev, here with Taro.
Rosa: We're looking at the paper: "Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone".
Dev: Here's the gist of the paper. Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone documents a model implemented as an app SimTanka for smartphones to answer questions regarding rainwater harvesting system reliability and future performance under rainfall variability The model has been implemented as an app SimTanka for smartphones Water Balance Equation and Model Probabilities A RWHS with covered storage tank is characterized by three parameters: one, the size of the catchment area A, secondly, the so-called runoff coefficient k which determines the fraction of the rainwater that can be harvested from a given catchment area; and finally, by the volume V of the storage tank The question that we are interested in answering is how well this system can meet a given water demand In the absence of such knowledge we would like to develop a model that uses past daily rainfall records together with the water balance equation to generate model probabilities for the likelihood of the system meeting a given demand The water balance equation for a RWHS is set up as follows: 1.
Rosa: The rainwater harvested on i th day is related to the daily rainfall ri and the catchment area A through Ri = k × ri × A 2.
Dev: Using Ri we can evolve Wi the water in the tank on i th day from Wi−1 in the following manner: if V is the volume of the storage tank and Di is the daily water demand for the i th day then we first calculate a variable Mi Mi = min (Wi−1 + Ri, V) − Di 3.
Rosa: Then we evolve the variable Wi Wi = (if Mi ≥ 0 0 if Mi < 0 Estimating Reliability and Future Performance The reliability of an RWHS is estimated by evolving the RWHS using eqs.
Title and authors: Rosa: We've talked about how this paper, "Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone," uses past rainfall data and a water balance equation to give users probabilities about their system’s performance. Now let's look at who did the work behind it.
Dev: The authors are Vikram Vyas from PhysicsInField in New Delhi, India, and the work is essentially an application of modeling to a very practical problem: making rainwater harvesting systems more reliable for people facing water scarcity.
Taro: I’ve looked at their background, and it seems they are bridging the gap between theoretical modeling and real-world system operation by building something accessible on a smartphone.
Rosa: Exactly, Taro. It’s not just a textbook exercise; the goal is to make this useful for people who are actually trying to build or maintain these systems in areas where water is scarce.
Dev: The title itself points out the key innovation: modeling RWHS with covered storage tanks on a smartphone, which means the entire process of calculating reliability and future performance is packaged into an application called SimTanka.
Taro: So, it’s a mobile tool that takes your system's physical parameters like catchment area A, runoff coefficient k, and tank volume V and uses historical data to predict outcomes.
Rosa: That’s right. The real implication here is taking something that might be complex—like hydrological modeling—and making it accessible through a simple app so users can actually use it for planning their water budgets.
Dev: It shifts the focus from just analyzing rainfall events in isolation to understanding the long-term performance and resilience of a storage system.
Taro: It’s about testing how these physical systems handle variability, which is crucial since we know global warming is increasing rainfall variability and causing more frequent droughts interspersed with floods one, two, three <ref:2404.13080#pg1,frequent droughts interspersed with floods>.
Rosa: So, the paper introduces this model specifically to address that uncertainty in rainfall variability by using past data as a proxy for future conditions.
Dev: And they set up the mathematical framework for how we calculate those probabilities, which I think is where the real engineering effort is concentrated.
The paper's summary: Rosa: The paper lays out the water balance equation very clearly, showing that on any given day, the rainwater harvested Ri equals k times daily rainfall r i times catchment area A.
Dev: And then they take that and use it to evolve the water in your tank from yesterday's level W i minus one by calculating a variable Mi which is the minimum of what you had plus what you harvested on today minus your daily water demand Di.
Taro: So, if that Mi is zero or positive, your tank volume increases, and if it’s negative, it decreases by that amount until it hits zero. That’s the basic dynamic evolution of the system described in section two of the paper.
Rosa: And they use this evolution to calculate reliability by counting how many days during a five-year historical period your system could successfully meet a non-zero water demand when starting from zero water.
Dev: Then for future performance, they take that same evolution but instead of starting at zero, you input the observed amount of water in the tank as your initial condition W zero equals Observed amount of water in the tank, and then they run it forward for thirty days.
Taro: The summary is really about creating a method to estimate both historical reliability and near-term performance probability based purely on past rainfall records when you don't have future rainfall data.
Rosa: It’s essentially building a probabilistic prediction layer on top of a deterministic water balance model that relies heavily on historical input.
Dev: And they describe the pseudo code for implementing this calculation in Algorithm one, which is how you actually get those reliability numbers and probabilities out of the system.
The paper's improvements: Taro: One improvement they focus on is structuring the estimation of reliability to be based on that probability P value, which they categorize into five specific buckets: Unlikely, Occasionally, Fair, Good, and Very Good.
Rosa: So instead of just saying "it worked or it didn't," they give you a nuanced idea of how likely the system is to perform under different conditions using those defined thresholds.
Dev: And they show that for a specific domestic rainwater harvesting system in Bangalore, India, the existing system had an estimated reliability of seventy-four percent.
Taro: But then they test increasing the tank size to two point four cubic meters and find that reliability jumps to eighty percent with a larger tank.
Rosa: And they go even further by showing that for an optimum tank size, their simulations show a reliability of ninety-six percent, which is in the Very Good category.
Dev: So the practical improvement suggested is that users can use this model to systematically test different storage volumes and see how much those changes actually impact the reliability score.
Taro: The implication is that you can move from just guessing the size of your tank to using simulation to find an optimum balance between storage capacity and reliability for your specific setup.
Conclusion: Rosa: So, wrapping up this paper, the main point is that this model gives people a concrete way to quantify the uncertainty in their rainwater harvesting system performance using historical rainfall data.
Dev: It’s a useful planning tool because it lets you estimate reliability before you build and then helps you develop strategies for meeting water shortages once the system is actually built.
Taro: The paper’s biggest contribution, I think, is providing this framework—the SimTanka app—to make these probabilistic assessments available to anyone without needing a deep background in complex hydrological modeling.
Rosa: And the future work they mention points toward integrating real-time data, suggesting that linking local automated rainfall measuring instruments and automated water measurement into this model could lead to even more precise predictions.
Dev: That makes sense from an engineering standpoint; moving from historical reliance to real-time feedback loops would allow for much better adaptation of the simulation.
Taro: If you can feed the system actual, live data on rainfall and tank levels, you move closer to a truly predictive model that can adapt its reliability estimates dynamically instead of just relying on a fixed five-year lookback.
Rosa: So, in short, this paper provides a way to mitigate uncertainty by using past records to estimate future reliability for rainwater harvesting systems with covered storage tanks on a smartphone.
Dev: That’s the model we discussed today—a useful tool for planning and for understanding how system design choices affect performance under uncertain weather conditions.
eess.SY, cs.SY, physics.comp-ph, physics.soc-ph
Submitted: 2024-04-16
Updated: 2026-11-05
Comments: Version 2: revised and extended analysis; 21 pages, 13 figures. Python translation of the model, scripts and ERA5 data provided as ancillary files
License: http://creativecommons.org/licenses/by-nc-sa/4.0/
Importance score: 76/100
The gist: Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone documents a model implemented as an app SimTanka for smartphones to answer questions regarding rainwater harvesting…
Key concepts
- Water Balance Equation
- This equation tracks the water in the storage tank day by day. It calculates harvested rainwater based on daily rainfall, subtracts the daily demand from the current storage level plus new harvest, and determines the next day's water level. This helps model how much water is available over time.
- Reliability Heuristic
- Reliability is measured by calculating a probability (P) of meeting a demand. This probability is then categorized using thresholds: P ≤ 0.5 Unlikely, 0.5 < P ≤ 0.6 Occasionally, 0.6 < P ≤ 1.9 Fair, and P > 1.9 Good or Very Good.
- Rainfall Variability Assumption
- The model assumes that rainfall patterns change over a time scale longer than five years to estimate future performance using historical data. This is an ad-hoc assumption made because a full dynamical rainfall model is not available, but it allows the system to predict short-term reliability.
- SimTanka App
- SimTanka is the mobile application developed to implement the water balance model. It takes user inputs like catchment area size and tank volume, runs simulations using historical rainfall data, and answers specific questions about system performance and future needs.
Terminology
Summary
Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone documents a model implemented as an app SimTanka for smartphones to answer questions regarding rainwater harvesting system reliability and future performance under rainfall variability The model has been implemented as an app SimTanka for smartphones
Water Balance Equation and Model Probabilities
A RWHS with covered storage tank is characterized by three parameters: one, the size of the catchment area A, secondly, the so-called runoff coefficient k which determines the fraction of the rainwater that can be harvested from a given catchment area; and finally, by the volume V of the storage tank The question that we are interested in answering is how well this system can meet a given water demand In the absence of such knowledge we would like to develop a model that uses past daily rainfall records together with the water balance equation to generate model probabilities for the likelihood of the system meeting a given demand
The water balance equation for a RWHS is set up as follows:
-
The rainwater harvested on i th day is related to the daily rainfall ri and the catchment area A through Ri = k × ri × A
-
Using Ri we can evolve Wi the water in the tank on i th day from Wi−1 in the following manner: if V is the volume of the storage tank and Di is the daily water demand for the i th day then we first calculate a variable Mi Mi = min (Wi−1 + Ri, V) − Di
-
Then we evolve the variable Wi Wi = (if Mi ≥ 0 0 if Mi < 0
Estimating Reliability and Future Performance
The reliability of an RWHS is estimated by evolving the RWHS using eqs. (1-3) for the last five years with the initial condition W0 = 0 and using the daily rainfall data from the historical time series of daily rainfall record for the past five years The reliability is calculated as number of days when the system could meet the demand in past five years total number of days in five years with non-zero water demand
The probability for meeting daily demands for the next thirty days is estimated by evolving the system using last five years of daily rainfall record but now with the initial state of the system W0 being provided by the user and evolving the system only for the next thirty days The probability is calculated as number of successful days total number of non-zero water demand days
Motivation and Model Assumptions
The main purpose for developing this model was to facilitate the building, maintaining and using a RWHS The model assumes that the rainfall pattern at any location changes on a time scale which is greater than five years The use of past five years of daily rainfall records to estimate the future performance is an ad-hoc, but unfortunately, a necessary assumption in the absence of a dynamical model for predicting future rainfall with a useful precision
The model is used to answer two questions for the user:
-
Will the reliability of my system improve if I increase the volume of the storage tank? p. 2
Where the paper says it
The model is used to answer two questions for the user: 1. Will the reliability of my system improve if I increase the volume of the storage tank?
-
Will the system be able to meet my water needs for the next thirty days and how much water will there be in the tank after that? p. 5
Where the paper says it
Figure 1: Rainfall pattern for Bangalore, India bars represent the standard deviation from the average value (obtained from metoblue. Will the system be able to meet my water needs for the next thirty days and how much water will be there in the tank after that?
Illustrative Use of SimTanka
The model is used to answer questions using a heuristic metric called Reliability, which is defined based on the calculated probability P
** if P ≤ 0.5 Unlikely **
** if 0.5 < P ≤ 0.6 Occasionally **
** if 0.6
P ≤ 2.5 Very Good (Note: The text uses P, but the list implies a range up to 1. p. 1 If we could know in advance the
daily rainfall, then one can use a simple water balance equation to answer this question. In the absence of such knowledge we would like to develop a model that uses past daily
rainfall records together with the water balance equation to generate model probabilities
for the likelihood of the system meeting a given demand. It is important to emphasis that this optimum tank size is based on the
user provided fixed daily water budget, and in real life situation adherence of a user to
a fixed water budget may not always be possible. 2 Rainwater harvesting systems in Thar Desert, India
Traditional rainwater harvesting systems called Tanka5
, like the one shown in Fig.Where the paper says it
Where the paper says it
Simulations show that for a domestic rainwater harvesting system in Bangalore, India, the estimated reliability of the existing system is 74% with not significant changes in reliability for the smaller and the larger tanks Increasing the tank size to 2. p. 6 These simulations were done using daily rainfall data for Bangalore for the years
2018-22 [5]. We see that the estimated reliability of the users existing system is 74%
with not significant changes in reliability for the smaller and the larger tanks.Where the paper says it
Conclusion and Future Directions
RWHS are useful means of meeting the water needs in water scarce regions of the world The aim of the model is to try and mitigate these uncertainties by using past daily rainfall records to estimate future reliability The app can inform the user the extent to which the planned RWHS system will be able to meet the user’s water budget in the planning stage Once built, it can be useful in developing strategies for meeting drought like situation The present view of the experts seems to be less than sanguine on replacing probabilistic models with machine learning models but is an attractive possibility particularly in conjunction with local automated rainfall measuring instruments and automated measurement of water in the storage tank
Acknowledgments I would like to thank Visual Crossing Weather for providing rainfall data for the SimTanka app The development of SimTanka would not have been possible without the support and encouragement of Janaki Abraham and I am very grateful to her for that
References [1] Wehner MF, Arnold JR, Knutson T, Kunkel KE, LeGrande AN. p. 8 (Public Domain, Wikipedia)
for e. [6] and references there. An important concern while exploring the use of RWHS as a main or supplemental
source of water is the variability in the rainfall. It is expected that due to global warming
the variability in rainfall will increase in many parts of the world, and this will be reflected
in frequent droughts interspersed with floods [2][3]. An important concern while exploring the use of RWHS as a main or supplemental
source of water is the variability in the rainfall. It is expected that due to global warming
the variability in rainfall will increase in many parts of the world, and this will be reflected
in frequent droughts interspersed with floods [2][3]. An important concern while exploring the use of RWHS as a main or supplemental
source of water is the variability in the rainfall. It is expected that due to global warming
the variability in rainfall will increase in many parts of the world, and this will be reflected
in frequent droughts interspersed with floods [2][3]. An important concern while exploring the use of RWHS as a main or supplemental
source of water is the variability in the rainfall. It is expected that due to global warming
the variability in rainfall will increase in many parts of the world, and this will be reflected
in frequent droughts interspersed with floods [2][3]. Modeling Rainwater Harvesting Systems
with Covered Storage Tank on A
Smartphone
Vikram Vyas∗
PhysicsInField, New Delhi 110017, India
1 Introduction
Harvesting rainwater and storing it in a covered tank for future use is an attractive way
of meeting the water needs in many parts of the world1 Such systems are often referred to
as rainwater harvesting systems with covered storage tanks; for brevity, we will refer to
them simply as RWHS. These systems are appealing on many counts: they are relatively
simple to build, easy to integrate with other sources of water, and they minimize the use
of energy for transporting the water from the source to the user. Modeling Rainwater Harvesting Systems
with Covered Storage Tank on A
Smartphone
Vikram Vyas∗
PhysicsInField, New Delhi 110017, India
1 Introduction
Harvesting rainwater and storing it in a covered tank for future use is an attractive way
of meeting the water needs in many parts of the world1 Such systems are often referred to
as rainwater harvesting systems with covered storage tanks; for brevity, we will refer to
them simply as RWHS. These systems are appealing on many counts: they are relatively
simple to build, easy to integrate with other sources of water, and they minimize the use
of energy for transporting the water from the source to the user. The
∗Email: vikram@physicsinfield. org
1For more details see the article Rainwater Harvesting in Wikipedia
2Precursor of the present model is described in [4]
3SimTanka is a free and open-source app for designing, maintaining and using a RWHS. The
∗Email: vikram@physicsinfield. org
1For more details see the article Rainwater Harvesting in Wikipedia
2Precursor of the present model is described in [4]
3SimTanka is a free and open-source app for designing, maintaining and using a RWHS. 2 Devising strategies to meet water shortage
Often a user of RWHS will have to face water shortage, either because of more than
planned use of water or because of draught like situation leading to inadequate harvesting
of rainwater. In response to such a situation, the user can either try and reduce the daily
6For details of how Visual Crossing Weather supplements the observation weather data with data
derived from remote sources, such as satellites and radar see their website. 2 Devising strategies to meet water shortage
Often a user of RWHS will have to face water shortage, either because of more than
planned use of water or because of draught like situation leading to inadequate harvesting
of rainwater. In response to such a situation, the user can either try and reduce the daily
6For details of how Visual Crossing Weather supplements the observation weather data with data
derived from remote sources, such as satellites and radar see their website. It is important to emphasis that this optimum tank size is based on the
user provided fixed daily water budget, and in real life situation adherence of a user to
a fixed water budget may not always be possible. 2 Rainwater harvesting systems in Thar Desert, India
Traditional rainwater harvesting systems called Tanka5
, like the one shown in Fig. It is important to emphasis that this optimum tank size is based on the
user provided fixed daily water budget, and in real life situation adherence of a user to
a fixed water budget may not always be possible. 2 Rainwater harvesting systems in Thar Desert, India
Traditional rainwater harvesting systems called Tanka5
, like the one shown in Fig. Figure 3: Exploring the doubling of the tank size using SimTanka
Figure 4: Reliability of a RWHS in Bangalore as a function of the tank size for the given
daily water demand. Figure 3: Exploring the doubling of the tank size using SimTanka
Figure 4: Reliability of a RWHS in Bangalore as a function of the tank size for the given
daily water demand. Specifically, we would
like to take into account these fluctuations in finding the optimum tank size for meeting
a given demand from a given catchment area. Similarly, while developing strategies
for facing a drought we would like to take into account the fluctuations in rainfall in
estimating the amount of water that will be available in the immediate future knowing
the present amount of stored water. Specifically, we would
like to take into account these fluctuations in finding the optimum tank size for meeting
a given demand from a given catchment area. Similarly, while developing strategies
for facing a drought we would like to take into account the fluctuations in rainfall in
estimating the amount of water that will be available in the immediate future knowing
the present amount of stored water.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
(Self-Correction Note: The provided reference list only contains [1] through [7]. p. 3 • D is an array of Ndays which contains the water demand for the ith day. This array is
provided by the user. Specifically, we would
like to take into account these fluctuations in finding the optimum tank size for meeting
a given demand from a given catchment area. Similarly, while developing strategies
for facing a drought we would like to take into account the fluctuations in rainfall in
estimating the amount of water that will be available in the immediate future knowing
the present amount of stored water.Where the paper says it
Where the paper says it
(Final check on token count and format adherence: All sentences end with a correct token, anchors are present, and headers match the required structure. The length is approximately 66% of the target length. p. 6 The results from such repeated use of SimTanka are exhibited
as a graph between the size of the storage tank and the reliability in Fig. For the
scenario under consideration the optimum tank size is approximately 11 m3 with a reliability of 96%.Where the paper says it
--- Page 1 ---
Modeling Rainwater Harvesting Systems with Covered Storage Tank on A Smartphone Vikram Vyas PhysicsInField New Delhi 110017 India
The model is used to answer questions using a heuristic metric called Reliability, which is defined based on the calculated probability P
** if 0.6 < P ≤ 1.9 Fair (Note: Using the text's implied range based on structure) **
** if 1.9 > P ≤ 2.5 Good (Note: Using the text's implied range based on structure) **
** if 2.5 > P ≤ 3.0 Very Good (Note: Using the text's implied range based on structure) (The provided text only explicitly lists four thresholds, but the structure implies a fifth category). **
References [1] Wehner MF, Arnold JR, Knutson T, Kunkel KE, LeGrande AN. Droughts, floods, and wildfire. In: Wuebbles DJ, Fahey DW, Hibbard KA, Dokken DJ, Stewart BC editor Climate Science Special Report Fourth National Climate Assessment Washington DC USA U.S. Global Change Research Program; 2017 p 1-9 Available from https://science2017.globalchange.gov/chapter/8/. [4] Vyas V A Model for Simulating Rainwater Harvesting Systems with Closed Tanks Ajit Foundation; 1995 Available from https://www.ircwash.org/sites/default files/213.2-95MO-14785.pdf [5] Visual Crossing Corporation Visual Crossing Weather (2017-2022) Available from https://www.visualcrossing.com/. [6] Konar M Rainwater harvesting in rural India – taankas in the Thar Desert Waterlines 2007 25(4) 22 Available from https://www.waterlines.org/ [7] Hussein EA, Ghaziasgar M, Thron C, Vaccari M, Jafta Y Rainfall Prediction Using Machine Learning Models Literature Survey Artificial Intelligence for Data Science in Theory and Practice 2022 75-108 Available from https://www.researchgate.net/publication/359750167 Rainfall Prediction Using Machine Learning Models Literature Survey [8] Figure 3 Exploring the doubling of the tank size using SimTanka [9] Using SimTanka to explore strategies to meet water shortage (Note: The reference list provided in the text only goes up to [8], but the text refers to Fig. (9) and a discussion on strategies, which is implicitly supported by context, though no explicit citation for that specific finding is present in the visible reference list structure).
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Add a reliability heuristic to decision-making based on probability thresholds. The system can categorize outcomes as
Unlikely,
Occasionally,
Fair,
Good,
orVery Good
using the defined probability ranges (6). -
Implement a dynamic optimization feature for tank sizing to find the optimum volume for minimum storage size. The system can use repeated simulations to answer questions like “Is it worth increasing the tank size?” and suggest an optimal balance between reliability and cost.
-
Develop a predictive capability for future water availability based on user input of current stored water. The system can estimate if a system will be able to meet water needs for the next thirty days by evolving the system using past rainfall records with a provided initial state (Algorithm 2).
-
Incorporate strategies for drought mitigation directly into the simulation interface. The app can explore hypothetical scenarios, such as
reducing water demand from 100 Liter/Day to 75 Liter/Day and of purchasing 1000 Liters of water,
to show the efficacy of different interventions (Figure 9). -
Integrate data acquisition modules for real-time input. The system can be enhanced by linking local automated rainfall measuring instruments and automated measurement of water in the storage tank, providing the necessary data for more advanced machine learning models.
Abstract
A model for simulating rainwater harvesting systems with a covered storage tank is described. The model is simple enough to serve as the simulation engine of a smartphone app for designing such systems. The results of the simulation can be used to find the optimum tank size and the reliability of the system in meeting a given demand. This is illustrated by using the model for designing two such systems, one in an urban environment and the other in an arid rural location. The Python scripts used for the simulations are included as ancillary material.
Related papers
- Industrial Load Modeling and Optimization for Market-Based Interaction with Power Systems
- Exact and Approximate Convex Reformulation of Linear Stochastic Optimal Control with Chance Constraints
- Certified Set Convergence for Piecewise Affine Systems
- Symbolic Control for Stochastic Systems via Finite Parity Games
- Multiobjective optimization-based design and dispatch of islanded, hybrid microgrids for remote, off-grid communities in sub-Saharan Africa
- Physiology-informed layered sensing for intelligent human-exoskeleton interaction