Optimization of Refrigeration Systems and Process Interactions
With more than 500 projects completed with less than one year to payback, Cargill Optimizing Services understands how to decrease costs and increase yields in the food and beverage industry.
This article reviews eight potential changes to a “simple” two-stage, ammonia-based refrigeration system, and investigates both the impact of individual changes and their possible interactions. The best possible savings of 25% operating costs can be achieved if the process interaction is considered. The conclusion, for this test system at least, is that along with floating head pressure, it is the interactions with the refrigeration system’s demand (the operating plant) that have the biggest potential to save operating cost. Thus a holistic site model has the greatest potential to save energy in the freezer room and/or increase the capacity of the plant.
- Asset Utilization
Two-stage refrigeration systems operate in a closed recycle and respond in a highly nonlinear fashion. Close coupling to process conditions introduces additional interactions and further complicates the prediction of performance. In such a complex and interactive system, opportunities abound to increase capacity / reduce cost – as well as looking at the refrigeration system in isolation, the interaction between the utilities and the process can be a fruitful area of investigation. Several typical options are investigated in this article and found to deliver savings of up to 25%. While potentially interesting from an energy cost perspective, even relatively small savings can become much more attractive if freezing capacity is a bottleneck.
In this study, a two-stage, ammonia system was modeled. The performance of the two screw compressors was based on design information that characterized the impact of suction and discharge pressure and slide valve position on capacity and motor load. Compressor oil was cooled externally with water. The evaporative condenser (EC) captured airflow, water evaporation and recycle and the load on the fans. The plant’s demands on the refrigeration system were treated as point loads of 225 tons of refrigeration (TR, 2.85 GJ/h) at -40°F (-40°C) and 150 TR (1.9 GJ/h) at 30°F (-1°C).
Eight modifications were reviewed individually and in combination and some sensitivity into the impact of ambient conditions was performed:
|Float head pressure||Desuperhead low stage||Shift load (low to high)|
|Intercooler pressure||Desuperhead high stage||Increase low stage temp|
Trade evaporator condenser
fan vs. high stage load
|Subcool HP liquid|
Floating the high stage head pressure was clearly the cheapest and most effective generally applicable change – this is a very common practice, and in this system, dropping the discharge temperature by 10°F (5.6°C) reduced electrical load by 5%. As the ambient temperature dropped, so the potential increased – a drop from 60 to 40°F (15.6 to 4.4°C) increased savings to al-most 10% - but relative humidity has much less impact. Within a reasonable operating margin, changes to the intercooler pressure had relatively little impact, even in combination with changes in high stage pressure. Similarly running the EC fans harder had relatively little impact in combination with the other operational changes – the additional fan load offset, and ultimately exceeded, the high stage compressor savings over the range reviewed.
All the capital changes have the potential to warm cold water entering the plant, and so could offset heating bills in addition to reducing electrical loads. All were considered alone and in combination with operational changes – changes in head pressure and (to a lesser extent) intercooler and EC fan speed made the changes much more attractive. Subcooling the high pressure liquid from 100 to 60°F (37.8 to 15.6°C) was the most promising – it can reduce electrical load up to 8%. Cooling the low stage vapors from 140 to 70°F (60 to 21.1°C) reduces the overall electrical load by 7.3%, while desuperheating the high stage vapors offers 6.5%. As is typical, these benefits are not additive – in combination, they “steal” from each other – combing all three capital options offered up to 11.6% reduction in cost.
Perhaps the greatest potential comes from changing the demand side of the equation. Shifting just 25 TR (0.32 GJ/h) from the low side to the high side compressors has the potential to save 9.4% of the total electrical load, while raising the low side to -30°F from -40°F (-34.4°C from -40°C) can save 12.5%. Combining these two options have some interaction – together they offer up to 16% of electrical usage.
Combining the capital, process and operational changes offers 22% savings potential when ambient conditions are 60°F (15.6°C) and 60%RH – this rises to 25% if the ambient temperature drops to 40°F (4.4°C).
Clearly, the interaction between the process and the utilities is a lucrative area to review, and a detailed understanding of process-side conditions and potential modifications in that area can have very significant impacts.
Caveat: This case study is based on a relatively simple implementation of a 2-stage system, and the results achieved are a very strong function of the initial conditions chosen. Each system in the field will be configured differently, with different compressor curves and process-side demands. In the real world, systems are often more complex that in this simple example – for example, multiple compressors may exist, and their unloading profiles can be optimized; process loads vary; evaporators may require defrosting; elec-trical rates may vary over the course of the day.
These results are intended to demonstrate that there are several potential operational and capital modifications that can be performed, and that those modifications interact significantly in the highly integrated and non-linear world of refrigeration system performance and system interaction with the process. Finding your way through the tangle of options and optimizing day-to-day performance is greatly enhanced with a process simulation.
When to consider
The value of a refrigeration system model is proportional to the complexity of the sys-tem. The greater the swings in process demand and am-bient conditions, the more individual compressor units and pressure levels are han-dled, and the more points of process interaction, the greater the potential for impact. Investigations can be particularly valuable if freezer capacity is a constraint on production, if capital invest-ment is planned in the process or engine room, or if significant shifts in produc-tion are planned.
About the Author
Charles Sanderson is the Technical Director of CPO, based in Minnesota, USA. He has led the development of Cargill's simulation capa-bilities for over ten years. He and the team have deployed the tools in projects across Europe, Asia, Africa and the Americas. Charles graduat-ed in Chemical Engineering from Imperial College, Lon-don and received his PhD from Sydney University in Australia.
With more than 500 projects completed that have less than one year to payback, Cargill Optimizing Services understands how to decrease costs and increase yields in the food and beverage industry.