Industrial Automation Magazine India

Predictive Maintenance & Pharma Manufacturing

Published on : Tuesday 12-05-2020

These tough times demand technological innovation and the need to adopt the best possible technology available, says Monak Modi.

We are in this together! We have heard this a lot in past 3-4 months. The world economies have shrunk a lot in these months! Pandemic is the word we have heard more than any other. People need to believe that we can survive this. The only sector that is currently flourishing is Pharmaceutical manufacturing industries. Talking about India Pharma, the company’s shares are skyrocketing and India has already started to cash in. India has been a supplying lot of medical tablets to many developed nations due to its technical supremacy in medical science. Every multinational is currently operating from home. However, when this lockdown worldwide starts to get some relaxation, most of these companies would again occupy these rental spaces. The commercial buildings will again be booming with the human capital and will again start harnessing natural capital. People will be more cautious about hygiene and the thermal comfort whenever normalcy restores. Under any circumstances, air quality and thermal comfort are the most critical aspects of exceptionally performing Commercial buildings. Since social distancing has been the mantra to tackle Covid-19 it is imperative that most of these buildings will be managed with very less work force. It is going to be a challenge in terms of maintaining operational efficiency and safety but that is when the technological innovations would take front seat. The technology is advancing at a rapid pace and the HVAC & Refrigeration industry is no different.

Chillers/chilled water systems are at the centre of both these industries. And the quality of output in terms of pharmaceutical tablets (for pharma manufacturing) and the work productivity (for commercial buildings) by and large depends upon the how you operate the utility systems. Energy efficiency and continuous operation without any downtime are the prime objectives of the maintenance team at site place. The energy guzzlers in the both of above-mentioned cases are HVAC systems. Major components of these HVAC includes AHUs, cooling towers, compressor motors, ducts, etc. These components are prone to deterioration as the time goes and operational period grows. It’s important to have an actionable insights based on the data available through various sources to keep this component at their full potential as people rightly say data is the new oil. So let us get into some technicalities:

Chillers vs direct refrigerant-based cooling system

The output from chiller is used to reduce the temperature and generate more or less moisture free air in commercial buildings, pharma formulation divisions and institutional facilities. Chillers are primarily used to generate chilled water. Chillers are favoured against direct refrigerant-based cooling systems since they have low cost of ownership, proposing cost savings to the possessors of high-performance buildings across their lifespan. Properly engineered chillers work with HVAC systems to provide the right temperatures, humidity levels, and ventilation for the space, while also prioritising low operating cost and energy efficiency as well as promising low sound levels and nominal environmental impact. Another advantage is that chillers contribute to the sustainability of high-performance buildings because they operate at lower head pressures, which uses much less energy.

In industrial environment chilled water or other liquid from the chiller is pushed through process or laboratory equipment. Industrial chillers are used for controlled cooling of products, mechanisms and factory machinery in a wide range of industries. They are often used in the injection mouldings, metal processing, weld processes, machine tool industries and die casting furnaces/machines, pharmaceutical formulation, F&B processing, thermal power stations and gas cooling in various applications.

The evolution of chillers through the years

Even though the elementary sections of a typical refrigeration systems and fluid chillers have remained untouched for decades, enhancements in compressor technology, refrigerant gases, manufacturing methods, and PLC/microcontroller based controls have transformed chillers into futuristic equipment with increasing functions and applications in not only comfort applications, but also in the industrial sector. Integrated circuit technology, Building Management Systems (BMS) communication, and data analytics have become powerful and are used to control chiller systems more efficiently than previously possible. These days, a single chiller control module can have multiple inputs/outputs and links via Modbus, BACnet, and other BMS communication protocols. Basic chiller controls now include proportional integral derivative (PID) algorithms that read information during the operation of the chiller, monitor system parameters, and then control various functions, such as fans, compressor loading, and expansion valves, to achieve the required set point, and do so without overshoot or cycling on and off, thus saving energy.

One of the biggest areas of development in chillers has occurred in terms of controls. Chillers have become far more suited towards part load conditions. In reality, most applications for chillers demand them spending less than 1% of the time operating at full load. In fact, VSDs are on everyone’s list of notable new technology in terms of chillers. A multipurpose unit is actually sized to cover the maximum peak demand, which, depending on the country where it is installed, may be in summer or in winter. The presence of variable speed drive technology allows units to effectively follow each combination of thermal loads required by the system, achieving total efficiency ratio (TER) values (calculated considering the unit’s performance in hot and cold water production) of up to 19% higher compared to traditional fixed-speed units.

The new age connected chiller

Additional path breaking innovation in controls is the ability to connect the chiller to the ‘cloud’ or remote servers. Data is sent from the chiller to the cloud where it is analysed, which empowers reporting to be generated on how the chiller is performing or if there are probable concerns. Data issued from equipment to the cloud, digital twin models, and remote monitoring are just a few of the technology trends that are changing how chillers are being serviced and maintained. With new acumens into efficiency and admission to the expertise of skilled technicians, today’s Smart Connected Chillers offer benefits including fewer disruptions, reduced downtime, and longer equipment life. User Interface and the visualisation graphs can drive Improvements in Building Performance. Remember the old adage that a picture is worth a thousand words? That approach, used with data from connected chillers, takes multifold information and presents it in a visual way, so critical knowledge is instantaneously conveyed. A dashboard provides a granular outlook on the data on how an equipment is performing. Today, the use of simple visual cues – like the use of red to highlight a problem – and increased customisation of interfaces mean chiller data is no longer represented by numbers and spreadsheets, but offered through rich, interactive graphics.

Chiller condition monitoring using cloud has many advantages. 

Chiller Analytics is based on chiller electrical parameters and refrigerant parameters obtained from the chiller controller. With these data and meta-data from site, Chiller analytics engines can detect major health related issues of HVAC Components such as:
a. Fouling of Heat exchangers – Condenser/Evaporator
b. Flow Deviations – Cooling Water/Chilled Water
c. Oil Filter clogging problems, and
d. Refrigerant Level/Leakage problems.

Chiller performance parameters (a few mentioned below) can also be calculated and tracked on a live basis and alerts for high/low limits can be set so that, possible corrective action be taken and energy wastage can be avoided.
a. Cooling Load delivered (TR)
b. Input Power/unit cooling energy delivered (kW/TR), and
c. Coefficient of Performance (COP).

For a pharmaceutical formulation processes where tablets are being made, these indicators become very critical for operations as deviation of 0.5-1% in these parameters can lead to the loss of an entire batch of medicines which any organisation don’t want in these tough times with pressure of running these plants for 24 hours. First, it eliminates manual logbooks, which uses lot of time and effort. Most importantly since the data is available remotely maintenance engineers team, OEM and energy technical services team, can suggest necessary actions. Finally, Machine learning can be incorporated into the analysis, which offers the operator with recommendations. This makes maintenance and uptime easier and delivers more predictability.

Predictive Maintenance of chillers
How can chiller analytics evolve more? There has to be 100% accuracy of data transmission between IoT Gateway and Cloud with appropriate access level. Chiller Analytics involve acquiring data from local controllers and sending live data to cloud. Conditions can be set on critical parameters, which can trigger SMS or emails to responsible stakeholders. The thresholds and violation limits can get mature over a time with every possible load conditions get observed. OEMs having access to this data can instantly look at data trends remotely and identify suitable corrective actions. The maintenance activities can also be planned by OEM to avoid any unplanned breakdown. OEMs can also suggest user to keep an inventory of

spares. So that inventory cost can also be optimised. Status of critical components within the chiller can be tracked and deterioration can trigger actions like spare procurement by end- user before breakdown. This will eventually lead to predictive maintenance of chillers. Condition monitoring of chillers can help to track key performance indices on a continuous basis. Usually, during a breakdown of chiller, the service personnel reads through the chiller logs maintained by the operator and tries to identify specifics patterns in data recorded to pin point what possibly could have gone wrong while the chiller was operating. Chiller predictive maintenance is a similar approach wherein the service personnel uses his knowledge to train the AI to identify specific patterns in data recorded. Chiller condition monitoring digitises the age-old logs maintained by operators, thereby enabling the AI instance running on a remote cloud to work on this data. Identifying specific patterns with volumes of data available is tedious if done manually, which is being solved with machine learning.

The future of chillers is smart

The future of chillers will see more developments into chiller predictive maintenance, where the OEMs will be notified on the events and proactively solve problems before the chiller breaks down. These tough times demand technological innovation and the need to adopt the best possible technology available. For pharmaceutical, industries and buildings where HVAC systems are the most sought after operations, machine learning algorithms can save lot of money. One thing needs to be understood there are no perfect algorithms/models. We need to be adaptive in nature and these models can get mature with the time and the quantum of data available.

Monak Modi looks after Industry Solutions at Utthunga Technologies. Solving business problems with the use of technologies for manufacturing industries is his passion. With rich background of Instrumentation and Energy Management, his goal is to create smart factories with sustainable initiatives and change management. He is always curious to be involved in new technologies that can heal the earth we live in.