Eye on Failure is a series of articles highlighting the stories of maintenance workers who have benefited from TRACTIAN by eliminating failures and unplanned machine downtime. To see other cases such as Embraer, Yara International, and AmstedMaxion, visit our blog.
Founded in August 1952, Sao Domingos Sugarcane Mill produces VHP brown sugar that’s destined for the global market. They also produce two types of ethanol – anhydrous is an additive in gasoline, and hydrous is used in flex fuel cars – yeast, electric energy, sugarcane fiber materials, and fuel oil.
Although the company’s been around for some time, Sao Domingos has always sought to be a pioneer in the industry. They achieve this by continuously applying cutting-edge technology for optimized, improved asset management.
Recently, Sao Domingos Sugarcane Mill has been undergoing a lot of growth. Their goal is to double their sugarcane production to 3 million tons, which will enable the company to significantly increase its sugar and ethanol production.
Prescriptive Maintenance Project Goals
The mill’s Maintenance Planning Coordinator led the project to implement TRACTIAN prescriptive monitoring. Although he had worked with an online predictive approach previously, prescriptive maintenance was news to everyone.
The Innovation and Digital Transformation Manager at the facility highlights that the main motivation for choosing TRACTIAN was the ability to solve operational problems, achieving a more efficient return on investment, and the ability to anticipate failures.
“We have a pretty big maintenance process, but since our manufacturing plant has been here for a while, the assets are old; even with up-to-date maintenance, anticipating failures has always been essential, along with raising assertiveness based on data and improving the availability of critical assets throughout the harvest period.”
An important aspect of the maintenance process at Sao Domingos Sugarcane Mill is that the team is small, yet capable. Currently, the Maintenance Planning department has six people, who are responsible for all maintenance scheduling and also for TRACTIAN’s online monitoring and fault prescription project.
Start of Project
Implementing online prescriptive monitoring at Sao Domingos Sugarcane Mill began in April 2022. First, the team selected in which machines the sensors would be installed, defined the project schedule, and the delivery of the sensors was carried out. TRACTIAN’s customer support team also provided training to start the project.
The whole journey was marked by high expectations of return on investment, both from the maintenance team and the company’s directors and operational decision-makers. This is because the average lost profit for an hour of line downtime varies between $3,053 and $4,071. In addition, the availability target set for 2022 was 98%, compared to 2021 when it was 96.4%.
The Artificial Intelligence-assisted prescriptive online monitoring was applied in all critical machines, such as cooling towers, pumps, reducers, and boiler exhaust fans. The main goal was to ensure the safety of employees and increase the amount of information and data available for decision-making. Currently, more than 70 critical machines are monitored in real time by TRACTIAN sensors.
The Maintenance Planning Coordinator said that the fault prescription alerts are followed up by the director. This allows them to intervene in the machines at the right time, and by scheduling tasks, they can operate without too many interruptions. According to him,
“We are managing to schedule and perform maintenance with excellence.”
Avoided Failures
One of the many failures avoided at the power plant was in the exhaust fan bearing of Boiler 1500, which works 24/7 to meet demands. Because it’s critical to the entire operation, if the exhaust fan stops, the whole boiler needs to be shut down.
The Innovation and Digital Transformation Manager reports that even though maintenance was performed during the off-season, when the harvest period started the monitoring platform’s Artificial Intelligence started to indicate an initial stage of wear on the external race of the exhaust fan bearing. This alerted the entire team, letting them know that if the failure progressed, the equipment would need to be shut down immediately.
By observing the spectrum presented in the platform, it was clear that the amplitude of the frequencies was increasing – meaning the equipment was reaching a critical point. That’s when the maintenance team scheduled a two-hour planned downtime intervention. All the necessary preparations were made, including selecting the responsible team and allocating the necessary resources.
Another fault prescribed by TRACTIAN Artificial Intelligence Platform was an unbalance in the BMP boiler, which has the highest flow rate in the plant.
The Maintenance Planning Coordinator says the platform prescribed a random unbalance fault. Because this asset in particular was highly critical to the operation, as soon as they received the alert, the failure was analyzed and brought to management. They then decided to stop the machine, and the unbalance was verified by dirt build-up on the rotor.
The planned downtime lasted for only two hours, just enough time to clean the equipment. After the repair, vibration levels returned to normal.
On a different case, TRACTIAN AI detected another critical failure in Cooling Tower 1, which is responsible for cooling the water supply for the distillery’s industrial park. This time, it was a temperature increase alert. The team mentioned that the failure seemed random, as the equipment was operating within normal standards before the alert.
As soon as the director received the prescriptive alert on his cell phone, he triggered the maintenance team to perform on-site inspection and clean the cooling tower motor fan. Once the inspection was finished, temperature decreased.
A fourth critical looseness fault was prescribed on Exhaust Fan 02 of Boiler 1800, which had been undergoing monitoring for a while. The maintenance team received the prescriptive alert of a gap in the key of the motor pulley that is coupled to the bearing, and they followed this path to the fault.
When they saw that the vibration patterns were changing, they scheduled all the necessary inspection and repair activities. When the maintenance team went to work on the equipment, they saw that the key was indeed loose. Additionally, they took the opportunity to open the bearings, change the grease, and check the tensioning of the belts.
The Change Generated by Online Monitoring
The Maintenance Coordinator stated that by the end of 2022, Sao Domingos Sugarcane Mill had very high availability rates for TRACTIAN-monitored equipment, exceeding the target set at the start of the harvest period.
“In the not-so-distant past, everyone thought maintenance synonymous with extra costs. I’m fascinated by good practices, technology and excellence, and we went after TRACTIAN solutions to keep production running with excellence, without machine failures or unplanned downtime.”
Sao Domingos Sugarcane Mill has a long production chain, meaning their capital needs to be well invested and show positive results for the company.
So, instead of suffering unexpected downtime and stopping production during the harvest period, the company can now intervene at the right time. They can schedule maintenance interventions so they have the smallest impact possible on production, and finally achieve the established production of sugar and ethanol goals defined at the beginning of the harvest.
If you like this approach and want to provide your industry with a prescriptive online monitoring solution – predicting failures and increasing the life of machines – contact one of our experts and get a demo.
