Medium temperature thermal energy storagefor high efficiency solar cooling applications

Pintaldi, S 2017, Medium temperature thermal energy storagefor high efficiency solar cooling applications, Doctor of Philosophy (PhD), Engineering, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Pintaldi.pdf Thesis Click to show the corresponding preview/stream application/pdf; Bytes
Title Medium temperature thermal energy storagefor high efficiency solar cooling applications
Author(s) Pintaldi, S
Year 2017
Abstract This thesis analyses thermal energy storage options for medium-temperature high-efficiency solar cooling systems with absorption chillers, by exploring new thermal storage technologies and advanced control strategies. Considerable effort was spent on building modelling and experimental tools required to systematically evaluate the control strategy and storage material impact in a solar cooling plant. They are:

A detailed system level dynamical model of a solar air conditioning system has been developed in the software TRNSYS as a part of this study. The system level model integrates both sensible and latent heat storage models. In order to accurately simulate the phase change heat transfer, a detailed numerical model for a latent heat TES was developed and validated with experimental data from a test rig built as a part of the thesis work.

In a typical solar air conditioning system, the controller functions to manage the plant response according to load demand without any information on the expected availability of energy sources or load. A Model Predictive Controller (MPC) with knowledge of solar and load forecast information can improve the performance of these systems. A Genetic algorithm (GA) based predictive controller was built and integrated with a simplified solar cooling numerical model.

As part of the project were the design, the installation and the commissioning of a test rig at the CSIRO Energy Centre in Newcastle, Australia for evaluating high efficiency solar air conditioning operation. The test rig consisted of 28 m2 single axis tracking Fresnel concentrating collectors, 1500 L thermal storage system, 20 kW backup heater and a 10 kWc double stage absorption chiller. Instrumentation and controllers were installed as part of the test rig to evaluate various operational strategies for improving the operation of a solar air conditioning plant. System level models were calibrated and improved through measured data from the test rig and used for further investigation of control strategies. A fan heat rejection system was used during the tests to mimic the chiller heat load.

In order to verify the benefits of MPC in an operating system with TES, a developed control framework was integrated with the test rig data logging and controls software (LabVIEW). Successful MPC implementation required seamless integration and feeding of test rig data to the MPC model, consideration of component response times (e.g heater), limitations such as pump minimum and maximum flow rates. The MPC controller of the test rig was implemented in Python.

The work reported in this thesis has improved our understanding and has developed suggestions and methods for optimal operation of solar air conditioning plants with thermal storage systems. Main learnings and outputs can be summarised as below:

This thesis provides a generalised methodology for evaluating the suitability of sensible and latent heat materials as thermal storage media for solar air conditioning applications. Solar cooling system annual simulations have been carried out under different scenarios to estimate the energy saving benefit of two sensible (water and thermal oil), two phase change (solar salt, Aluminium tin alloy) storage media. Comparison of the results showed that despite latent heat storage exhibiting higher storage performance, the performance of the complete solar cooling system is lower than plants with sensible heat storage due to higher temperature operation of the collector and the indirect heat transfer related losses. High thermal conductivity phase change material did not perform better than sensible storage materials. Proper alignment of phase change temperature and storage temperature could help in improve the performance of latent heat storage plants.

Component level tests showed the collector operating with 32 % efficiency and with intermittent tracking related problems. Storage tank showed typical stratification and heat loss coefficient of 4.9 W/K, as per design specifications. Test results showed non optimal operation of chiller. The latter delivered an intermittent cooling power of 6 kW, whilst absorbing 10 kW of heat, resulting in a Coefficient of Performance (COP) < 1. Therefore complete solar plant tests were carry out with a fan heat rejection system. During the test period, the solar collector produced 22.6 kWh of heat, and 10 kWh of heat was rejected through the fan heat exchanger. As a result, more than 50 % of heat was lost due to thermal losses in the system. These results stress the importance of running preliminary test on single component performance, and using conservative performance data in the design phase. Nevertheless the test rig can use some simple control strategies and can be used for MPC evaluation.

Numerical simulations of GA-based MPC of a solar cooling plant model showed v the ability of the MPC to take advantage of the heat stored in the tank, and successfully reduce backup heater usage. Ten day simulations using typical weather data have been carried out to compare conventional or Rule-Based Control (RBC) and MPC. Auxiliary heat reduction with respect RBC is up to 15 %. Detailed analysis of the MPC performance showed that benefit is maximum over period of low solar availability.
The objective of MPC implementation in the test rig is to gather first hand operational experience in a real world scenario, as the performance of the controller is strongly dependent on the real plant operation including the dynamics of the components. A subsystem of the test rig was utilised to evaluate a RBC versus a MPC strategy. The objective of the controller was to maintain the thermal store at a set point temperature throughout the day with minimal electricity cost incurred for heating. A time of use electricity tariff structure was used with peak rates after 0200 pm.

Comparison of RBC and MPC controller responses showed the MPC tends to utilise part load settings of the heater and the pump rather than the ON – OFF used in RBC. However, limitations due to pump operating range and heater settings made the control signal implemented in the real plant different from the MPC model control output. It is possible to overcome these limitations by incorporating the response behaviour of the pump and the heater in the MPC model.

Before implementing the MPC in real operating plant, a numerical model of the real plant was utilised to evaluate the benefits of MPC. Rather than operating the pump and the heater in ON – OFF mode, the MPC controller operated the system at part load conditions and achieved a 14 % reduction in the cost function.

Tests have been carried out to evaluate the potential of MPC in reducing backup energy cost with a-priori knowledge of solar radiation availability. The scope of the control logic was to charge the storage system to a set-point temperature at the end of the day with a time varying electricity cost. The results showed the ability of MPC to use forecast information related to radiation and electricity cost.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Engineering
Subjects Renewable Power and Energy Systems Engineering (excl. Solar Cells)
Control Systems, Robotics and Automation
Energy Generation, Conversion and Storage Engineering
Keyword(s) thermal energy storage
phase change material
solar cooling
absorption chiller
solar air conditioning
model predictive control
python
latent heat
optimization
experimental
Versions
Version Filter Type
Access Statistics: 257 Abstract Views, 378 File Downloads  -  Detailed Statistics
Created: Thu, 18 Jan 2018, 08:30:22 EST by Denise Paciocco
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us