Schuylkill Sensors

Guide Water recreational activities and raise public awareness about the water status of the rivers

Team: Saffron Livaccari, Jie Wang, Jiali Yao, Yebei Yao

Background

In 1940, the Delaware River was one of the most polluted in the nation. A dream of a clean river seemed terribly hopeless, as one EPA official stated in 1973 that the Delaware River may never be fishable (Kauffman Jr., 2010). Although hopes were low, attempts were made to clean the river. The Delaware River Basin Commission was created in 1961 to oversee the management of the river (About DRBC, 2022). The Clean Water Act (CWA) was passed by congress in 1972, which created the first federal regulations on sewage discharge into waterways (US EPA, 2013). Today, the Delaware River is finally safe to swim in again; except for 27 miles around Philadelphia, Camden, and Chester. This 27-mile stretch is only considered safe for secondary contact recreation – fishing or boating (Baillie, n.d.). The reason this area is unsafe for contact is in part due to bacterial contamination from combined sewer outfalls. The City of Philadelphia is attempting to assuage this persisting issue with the Green City, Clean Water program. This program is a 25-year plan to improve water quality through green stormwater infrastructure (Green City, Clean Waters: A Decade in the Community, 2021). While the city is attempting to ameliorate its issue of poor water quality, our project can complement it by helping inform and spread the concept of the quality of our rivers.

The Problem

The Delaware River and Schuylkill River in Philadelphia are designated unsafe for swimming. This is an equity issue, as many people in Philadelphia cannot vacation at the ocean or a lake to enjoy water activities in a safer environment. Many people will use the Delaware or Schuylkill River since there is no other choice. Therefore, our primary aim for this project is to inform the public about water quality and spark general public interest in the quality of our rivers.

The Project

Introduction

Our sensors will be partly submerged along the banks of the Schuylkill River upstream from Bartram’s Garden. The data will be aggregated to state the water quality at Bartram’s Gardens. Bartram’s Garden is a popular location for kayaking, fishing, and occasionally swimming. Therefore, we hope people will use this project to make a better-informed decision on the recreating they were already going to do. We will use pH level, turbidity, precipitation, and temperature to estimate the safety conditions for no contact, secondary contact (boating/fishing), and primary contact (swimming), which will be indicated by an OLED screen and LED lights at the docks of Bartram’s Gardens. Hopefully, we can help the public understand more about the quality of our rivers, inspire people to take care of our rivers, and strive for policies that do clean up the river.

Sensors and the system

The inputs will be: temperature, rainfall (water level in a cup), pH, and turbidity. And the outputs will be: an email flood warning based on if there is heavy rainfall; an LED light that turns yellow, red, and green; and a screen with turbidity, pH, temperature, and water contact safety. The pH, turbidity, and temperature will be displayed on the screen as well as images depicting the ability to swim, boat, or fish.

Water Quality Indicators	Safe Values
pH: Measure of chemical pollution	6.5 >= pH value <= 8.5
Turbidity: Measure of the clarity of the water	NTU <= 50

Turbidity is sensed by measuring the scattering of light from suspended solids in the water. The more suspended solids, the more turbid the water. Suspended solids could include organic matter, organic compounds, algae, clay, silt, etc (Turbidity and Water, 2018). Higher turbidity is associated with higher levels of bacteria, viruses, or parasites (US EPA, 2015). There are no general standards for turbidity; rather, some states have their own standards. Alaska has its own contact recreation standards which are less than 50 NTU (EPA, 2017). The pH of water is important to consider for water quality since changing pH can be an indicator of increasing pollution (PH and Water, 2019). A pH of 7 is neutral, and anything outside of 6.5-8.5 is unsafe (US EPA, 2015).

Prototype (click for the code)

A function, see the image below, within the code inputs pH and turbidity and outputs a safety score of 0, 1, or 2. A safety score of 0 means that it is safe to swim since the turbidity and the pH is within a safe contact range. A safe range of pH is between 6.5 and 8.5 and a safe range of turbidity is lower than 50. A safety score of 1 means that it is only safe to boat or fish, not to swim. A safety score of 1 would be output if either pH or turbidity is out of safe range. A safety score of 2 means the pH and the turbidity are both outside of their safe range and the water is not safe for any contact. These safety scores are input into a function that determines what is displayed on the screen near the shore.

If the water is not safe for contact:

If the water is only safe for secondary contact:

If the water is safe for all contact:

Our project includes a flood warning system as well. Flooding is determined by the speed of the rain; if rainfall is greater than 0.3 inches per hour (7mm/hour), then the flood warning will trigger. The significance of 0.3 in/hr is the qualification of ‘heavy rainfall’ according to the Meteorological Service of Canada (Barani, n.d.). If the sensor detects heavy rainfall, then an email notification will be sent out to whomever signed up for notifications and the screen near the shore will flash a warning sign. The screen will show the flood warning for a full 24 hours because of the possibility of delayed flooding and the fact that it is not safe to swim after a rainfall event.

If heavy rainfall is detected:

The Design

Since only parts of the sensors can be submerged in water, our sensors will have to be suspended above the river. We have designed a flotation device in order to accommodate our sensors:

Each Arduino sensor kit forms like the drawing on the left. The sensors will be tied to the shore via a rope and balanced above the river in a floatation system. The part of the sensors designed to not be submerged will be inside the cylinder and the part of the sensors designed to be submerged will be in the water at the bottom of the cylinder. The tube will be inflated for buoyancy. On top of the whole contraption, there will be one RGB light which will change between red, yellow, and green depending on the sensed water quality.

Interactive Display

While developing this project, we have considered four perspectives on people who may interact with our project.

In order to let everyone be involved in our project and learn more about the water around them. In addition to showing the water quality data on the screen, we also include different kinds of interactive displays on different sites.

Since the River runs linearly across the city and our devices should be adapted to different surroundings, we formalized three kinds of prototypes: which are for parks, trails, and bridges. In this way, the design concept can be scaled up more easily throughout the river.

Here, we choose three locations along the Schuylkill River to give more details about our designs.

Gardens and Parks

For a park located along the river, for example, Bartram’s Garden. To visualize the data collected by the sensor, we decided to place a real-time scrolling electronic screen in a conspicuous place, with a text description next to the screen to explain the factors affecting water quality and the purpose of our sensor project.

When the water quality is in good condition, the LED light at the top of the sensor kit is green, while the electronic screen will show the state of the river and the suitable water activities. When the light is yellow, it is only safe for indirect contact water activities like boating or fishing. When the light is red, it means not safe for any contact.

2. Linear Trails

For the very common and popular trails along the river, considering that their users are not all for fishing or swimming, others like pedestrians, runners, and bicyclists should also have the opportunity to perceive the existence and importance of water quality through our project. We have two strategies to make the water quality easy to sense and visualize.

First, by placing devices like a mist-maker or atomizer controlled by Arduino kits that can generate water mist along the walkway. When the water temperature exceeds the range during summer or hot days, the water mist will be diffused.

In addition, LED spotlights are placed at fixed points along the river, and when ph and turbidity values are out of the normal range, the spotlights light up and play the role of dyeing the river, giving people a visual impression that the river is polluted and amplifying visitors’ perception of bad water quality. And again, red indicates bad water quality, green indicates good water quality.

3. Bridge/Crossing

Last but not least, for crossings or bridges, like the new SCHUYLKILL CROSSING links between Grays Ferry Crescent and Bartram’s Mile under the old railroad crossing. We plan to place the sensors at each end of the crossing near the riverbank and connect the sensor to the water light board in the middle of the bridge. The board consists of a light panel embedded with water-sensitive LEDs. People walking by can sketch patterns or text on the graffiti screen. The water-light graffiti is time-sensitive, with the brightness of the screen dimming down as the water evaporates until it disappears. When the LED light at the top of the sensor flashes green, proving that the water quality is good and safe for direct contact, the board can be used, allowing people to pass by to understand the water quality in a more interactive way while freely playing with their sense of art.

Website and Email Notification

The information displayed on our screen will also be displayed on a website that people can access via a QR code near the screen. Our display at Bartram’s Garden will be in a clear plastic case and there will be a QR code that sends people to our website. The website will describe our project in more detail, with more information on the quality of the river, and the current water quality status. We also will have a website for people to get to know more about the water body that they are interested in as well as how our system works. The home page of the website will have the information on the status of the water quality, updated at 9:30 am, 2:30 pm, and 7:30 pm.

The other pages on the website will show how our sensors work, our evaluation system, how our devices are assembled, and water treatment; anything that would help the public get a better understanding of the whole broader project.

Our project can complement the Schuylkill River Cast. This website uses a model to predict water quality and state contact safety. This website uses Red, Yellow, and Green to indicate the predicted water quality, as our project has followed. The RiverCast only predicts between Flat Rock Dam (Manayunk) to Fairmount Dam (Boathouse Row). Therefore, this project can complete the water quality predictions for the remaining section of the Schuylkill River.

Map of the RiverCast zone. This location is much further north from Bartram’s Gardens (*Resources – Philly RiverCast*, n.d.).

For our email notification, we are using the Arduino Wifi Rev 2 which does not support Arduino IoT Cloud. So instead of the Arduino IoT Cloud, we are using the WIFININA library to upload our data to Google Firebase Realtime Database. And we can use the cloud messaging integration, for example, Integromat to send the command from Firebase Realtime Database to the IFTTT Email Service. For people interested in our project and the water quality of Sychukill River, they can scan the QR code or directly register from our website. They can get email notifications about the daily water quality report and flood warning messages.

Limitations

However, for future implementation and scaling up, there are still some limitations to overcome. On one hand, we want to keep the devices affordable; on the other hand, the low-cost sensors that we are using currently are not stable during the monitoring, so to have higher quality data, we would have to balance the cost and benefits and get the best sensors within the budget. To make the information more accessible, we would consider scaling this up with cellular instead of mobile hotspots. Even though we would anticipate the water status of the river would mostly be yellow, we have faith in the residents that this would encourage them to give more attention and action to protect the river rather than just be disappointed with the city.