Co-operative Distributed Home Energy Management Systems (HEAT Simulator)

Home Energy And Temperature (HEAT) simulator is a MATLAB/ Simulink based platform that has been developed at Advanced Diagnosis Automation and Control (ADAC) Lab. The simulator captures power consumption and heat dissipation of different residential appliances in order to simulate energy usage and temperature variation within a house. The objective of this independent study was to develop and build load control capability into the simulator. As part of the study, loads were classified into different categories based upon the flexibility with which they could be controlled. In addition, multiple-state power consumption models for loads were considered and built into the simulator since they represent load operation more closely. Mathematical models and heuristic control strategies for different load classes were then developed which allowed loads to be controlled in proactive and reactive manner. Reactive control scheme for individual rooms was then implemented in the HEAT simulator using StateFlow.

Advisor: Prof. Mo-Yuen Chow

Grid Network Mapping for Electrical Utility Distribution Systems

The electrical grid in most of the developing countries like Pakistan is largely unmapped. In case of power outage, it takes a long time to detect and correct the faults in the system. Moreover, it is very difficult to track down the entities responsible for power theft. We have proposed a method to map the electrical grid using the data from meters deployed in the region. Since most of the developing countries still don't have a high penetration of smart meters, our mapping algorithm would be applicable to both legacy and smart grids. We are also working to show how this kind of mapping can help to reduce the time in identifying the location of the fault, reduce the power theft and be useful in other applications such as phase load balancing.

Advisor: Prof. Nauman Ahmad Zaffar

Demand Response in High Stress Electrical Grids

Demand Response offers an effective way of mitigating cost and operational difficulties that are associated with peak demand in the grid. However high stress grids such as in India and Pakistan experience frequent outages making appliance scheduling quite a complex problem given the consumer preferences and comfort constraints have to be satisfied as well. Further these grids have certain load profiles (such as UPS) that are unique to them, which adds to the complexity. We plan to design an appliance scheduling algorithm that is robust & efficient and takes into account the peculiarities of these high stress grids. We also plan to explore alternative schemes that achieve the same goal of peak load reduction and cost minimization for such grids.

Advisor: Prof. Ihsan Ayyub Qazi

Distributed Generation Placement and Economic Dispatch in Smart Grids

With advances in smart grid the centralized generation facilities are giving way largely to distributed generation (DG) sources that are based on renewables such as solar PVs. The intermittent nature of these renewable energy sources makes renewable integration difficult. Thus the challenge is to integrate these sources at the most optimal points with in the grid to ensure a constant and economic supply of energy. This project leverages power flow and economic dispatch mechanisms, while also looking at other parameters such as land pricing and solar insolation (wind speed etc.), to identify the most optimal source of distributed generation as well as finding the most suitable points for distributed generation integration within the grid.

Advisor: Prof. Naveed Arshad & Prof. Nauman Ahmad Zaffar

Renewable Integration and Energy Optimization in Smart Buildings and Neighbourhoods

Undergraduate Thesis (IEEE Lahore Chapter - Top 20 Senior Projects)

The soaring global energy demands continue to fatigue the fossil fuel reserves. The situation calls for cutting down the energy costs and working towards a sustainable future. As a result, the renewable energy sources are becoming increasingly important. They present a viable long term solution for the energy requirements with little carbon footprint. Within the renewable realm, solar photovoltaic cells promise the greatest potential as a technology of choice with the widest range of possible output powers made even more feasible with lower production costs and improved efficiency. Still there are certain inherent issues of reliability, in terms of uninterrupted supply of electricity. These issues are addressed using supplementary sources of energy, for instance wind and biomass hybrid systems in feasible areas and diesel generators in others. In such hybrid setups, the generators have to be operated to cater to fluctuations in solar influx, night hours and peak energy requirements. A blind implementation of grid connected solar, which does not or poorly caters to the reliability constraints, can be quite inefficient. It does not shave off peak capacity of the grid, rather in times of no insolation, grid would still be required to produce the total energy requirement. Even then, production through solar cells will still reduce the fuel consumption when the source is present. However the overall high cost of implementation and increased operational costs incurred by the utility would trickle down to the consumers rendering the renewable integration unsuitable. A possible solution to this problem is to reduce the peak demand. This can be achieved by management of loads and electrical storage, keeping in view the consumption and generation patterns. An energy management comprises of both the software implementation and the control hardware.

• Download the Thesis [pdf]

Advisor: Prof. Nauman Ahmad Zaffar

Integration of Energy Optimization Solutions in Local Industries

The electrical grid in Pakistan has inefficiencies in different areas such as transmission/distribution, quality, reliability, protection, etc. These Inadequacies in asset protection, revenue leakage and energy theft constitute an overall troubled energy profile. The fixes usually proposed in this regard are directed at reducing distribution network losses and demand side management as applied to residential and commercial sectors. The lessons from the residential and commercial implementation of smart meters can be extended to industrial consumers to achieve better industrial efficiency goals and help relieve grid congestion. Our work reviews the smart meter technology and applications across various sectors. We point out the areas for power quality and energy efficiency improvement within the industries specifically in the context of Pakistan and propose ways to achieving them through smart meters. We have also shed light on the implementation dynamics to avoid the possible pitfalls that might render the solution useless.

• • Download the Technical Report [pdf]

  • Paper Published: Renewable and Sustainable Energy Reviews

Advisor: Prof. Nauman Ahmad Zaffar

Smart Lighting Energy Efficiency Solution for University Campus

For a university campus, lighting constitutes a significant portion of the total energy consumption. Therefore cutting down on the lighting consumption would mean large energy cost reduction. The feasibility study by LUMS Energy Optimization Initiative analyzed the use of smart energy efficient lighting based on LEDs. As part of the study, lights were subjected to harmonic analysis to check for any voltage and current transients. Further the light intensity was monitored to ascertain the light quality in comparison with the existing lighting. The results obtained were then used to propose recommendations to the smart lighting vendors for improvement in their lights. Given the lack of an effective regulatory authority in the country, which could determine and ensure the quality standards, this was particularly important. With the recommendations the vendors came up with improved lighting which were then installed across the campus, totaling about 42 kW. The installed lights are reported to have about 75% savings.

Advisor: Prof. Nauman Ahmad Zaffar

Energy Forecasting in Micro-grids

Distribution and consumer end energy management in smart grids is more viable when dealt with proactively rather than reactively. Knowledge of the forthcoming events can prove extremely helpful in shaving the load peaks and managing demands with in the grid. This measure avoids unnecessary strain on the grid infrastructure, ensures reliable electricity supply and enables energy saving. A micro-grid must be able to sustain its demands even when the power grid to which it is connected is not operating. Hence these proactive savings become even more pronounced in a micro-grid. So we worked on temperature based load forecasting in micro-grids using artificial neural networks and support vector machines. Hourly loads were forecast for 20 zones within a utility, using temperature data from 11 stations. The training data available was for about 4.5 years.

Advisor: Prof. Naveed Arshad