Emerging batteries and storage management strategies pave the way for microgrids

It may sound ludicrous, but the concept of producing electricity is not very different from producing vegetables. It’s a different thing that the quality of one is increasing, the other in grave pity.

We need more and more of electricity for our daily tasks – heating, cooling, cooking, commute (metro and electric vehicles); and so does the industry – refineries, paper mills, chemical/metal production, or even IT to power its mammoth data centres. But production time can differ from consumption time. The demand could be way below average production or 2-3 times of that. Not being able to store properly for later use has led to murkiest of consequences.

Electrical transmission networks across the world use electrical energy storage systems (ESSs) to get away with this sudden surge in demand, by storing when available and releasing when not. This not only reduces burden on utilities and device wear that occurs from high peak-demand, but saves costs by delaying grid reinforcements.

Distributed generation is also the next part of the grid puzzle countries are struggling to solve. Electricity production happens in bulk, mostly near fuel extraction centres to minimize transportation costs and loses. This has turned around with deeper proliferation of renewables.

With renewable energy production rising fastest of all, EESs have one more benefit – bridge the gap between power generation and its intended time of use to raise their reliability. Generation can now take place at/near the load centres, most of it derived from abundantly available sources like solar, wind or water head. When deployed in small scale network, termed microgrids, they not only ease the load on central grid and saves tremendous loses incurred in T&D, but creates a decentralized, independent and flexible system of production and distribution that is easy to initiate/shut-down, control, maintain cost effectively and above all, encourages the use of green power. But with limited production, it is important to ensure a regular and balanced supply is fed to consumers. Energy Storage Systems or ESSs play a vital role in strengthening the ground for autonomous distributed generation systems like these by eliminating a major roadblock – intermittency or unpredictability of renewable energy sources.

Hydrogen storage

The most widely used method of storing electricity with zero percent harm to environment is Pumped Storage.

Excess electricity pumps water to a height storing potential energy; to initiate backup, water is flown back down to run generators and return electricity. But with high space requirements, they are more suitable to grid level deployment, storing massive amounts of energy for long period of time. Hydrogen Fuel Cell also uses water as an exchange medium. Excess electricity is applied to convert it into its constituents Hydrogen and Oxygen. Electricity is regenerated from the recombination of these gases and process produces water as waste.

HFCs are modular storage units and can be scaled at a later stage using expansion units. Phosphoric acid fuel cells (PAFC) were the first commercially available hydrogen storage units. They use liquid phosphoric as electrolyte and platinum coated carbon paper as electrodes. Hydrogen itself is part of an interesting research that implements it as a fuel. With highest density of energy (chemical), about five time of coal, it is already being used to power fuel cell electric vehicles (FCEVs).

They currently have a round trip efficiency ranging 70%-80% and have a competitive cost compared to batteries. Apart from cryogenic liquid form and compressed gas, cheaper and safer methods to store Hydrogen are being developed for a more widespread use of HFCs. Physisorption/chemisorption techniques for material based storage keeps a more stable form at ambient temperature and pressures. Latest research on new materials like Zeolites (MCM-41) and nanostructured hydride materials like binary hydrides (MgH2, TiH2) aim to capture and store more Hydrogen, preferably in solid state. Tweaking the structure from bulk to thin film, nanoparticles and nanoconfined composites improve the hydrogen sorption properties, unlocking potential use in new technological applications. In the two above said cases, they are required to meet stringent gravimetric and volumetric adsorption requirement for storage capacity target set by DOE at 5% wt (1.8kWh/Kg).

Redox Flow Batteries 

A Nature article reported how Gallium Phosphide, a material largely deployed in PV panels can be used to create extremely tiny ‘nanowires’ to achieve production of photo-electrochemical Hydrogen directly from solar energy and water. Its yield is about one-fifth of light-electricity-fuel cell production, but it is in initial phases and also, way cheaper than the panel in comparison. Hypersolar is also developing a solar hydrogen generator that directly has an electrolyser integrated into a solar cell. Named the H2Generator, it is a single unit making Hydrogen directly from water under the Sun. Redox Flow Batteries Electric utilities largely depend on lead-acid batteries for their storage. Despite several upgrades, modifications (like the VLRA batteries) and a low energy-to-volume ratio they have significant market share and are forecasted to continue to be deployed at an increasing pace of 6.4% every year (2016 – Grand View Research, Inc)

Li-ion continues to grow at a similar pace. Having high energy density, lower maintenance and a solid state nature, they find perfect mate in renewables like solar. But its cost has remained a deterrent for bulk power storage and in developing countries its adoption has been anything but rapid.

A Flow battery or a Redox Flow battery draws similarities from a battery and a fuel cell. Two liquids (electrolyte) separated by a membrane circulate in their respective space with the dissolved components creating a potential for ion exchange that happens through the membrane.

Depending upon the type of electrolyte and its state, the membrane, the electrode, a Flow battery can be of various types. First developed in 1940s they are now being considered as a promising technology and market has seen a variety of commercially successful products, primarily using HydrogenLithium, Vanadium-Vanadium, IronChromium and Zinc-Bromide/ZincBromine chemistries among others. But the industry research push is to use organic materials to make them harmless and inexpensive, especially if needs to be integrated with solar generation systems. A research letter published in Nature dicussed the conceptualization of Redox Flow batteries that use table salt solution as electrolyte, simple dialysis membranes and electrodes made out of organic polymers.

At an estimated 10% cost of current RFBs, the letter cites a method to create metalfree, all-organic energy storage device fit for domestic or commercial use. The storage capacity of the prototype was adjusted to 10Ah per litre and it’s cycling stability tested for 10,000 repeated charge/discharge cycles at 20mA/cm2. It registered 80% retention in the capacity even in the static, un-pumped conditions and a stable voltage between 10%-90% state of charge.

Super Capacitors

Capacitors are used invariably in the power system to improve the quality of power supplied. They work by storing and releasing the charge, thereby changing the phase difference between the supply voltage and current, which in turn changes the power factor of the output electricity.

Compared to batteries, they can charge and discharge at a rapid pace as no chemical reaction is involved in the process. Electricity is stored as electrostatic charge on the surface of the material and can work for millions of charge/discharge cycles without degradation in capacity.

But they can store it for a very small amount of time. Depending upon the electrodes and the dielectric medium used, charge on capacitors may last from a nanosecond to a few hours. And the amount retained is proportional to the size. Therefore, generally a bank of capacitors is used to achieve the desired results even in a small setup.

Supercapacitors have charge storing capacities up to hundred times more usual (dielectric) capacitors. Several new approaches have been tried in the recent times for improving the electrochemical performance, charge retention at high discharge current and higher charge capacity per unit weight of the device. One such finding published in Sciencedirect used nitrogen-doped mersoporous carbon (OMC) electrodes to improve electrode wettability and chemical conductivity while providing additional pseudo-capacitance, Dan Liu et al. [2016].

It details large scale production of OMCs, in presence of amino acids as polymerization catalysts or nitrogen dopants, while maintaining highly ordered mesostructures. With a variable nitrogen content, flexible mesophase (3-D body centred cubic or 2-D hexagonal) and other techniques, the surface area can be significantly enhanced up to three times. In a symmetrical, two electrode configuration, it shows high capacitance of 186F/g (and 75% capacity retention at 20A/g current densitites) in ionic liquid electrolyte.

If clubbed with ESS requiring high startup time, supercapacitors can be used to decrease the overall response time of a storage system and hence find huge interest in related applications.

STORAGE MANAGEMENT

Using ESSs for a small system like rooftop or community might be easy to manage but zoom up things for an area level micro grid and things could become a little complicated to control. And storage systems are an expensive lot that can degrade easily if not sized and used correctly.

Various energy management strategies have been developed for meticulous planning of a power system incorporating renewable generation sources and storage. Starting from overcharging/ discharging avoidance, optimizing battery cost to automated control and Hardware in-loop simulations, several energy management scheme strategies have been developed to improve storage lifetime and operability. A couple of them are discussed below.

Receding Horizon control

RHC or Receding Horizon Control is a predictive method wherein a system can generate control variables to decide automatically upon the dynamics of charging and transmission taking into account the intermittency of the renewable sources. Using heaps of weather prediction data sets, modelling, load patterns and other related parameters, the system can be steered in a direction that makes it reliable and optimized.

The strategy takes advantage of prediction of the future generation in the power sources, the load requirement, and the evolution of the state of charge in the batteries. Predictions of the power sources and the load are performed using autoregressive models and historic data of wind speed, ambient temperature, solar radiation, and load demand. Thus, at a given time step, the energy management strategy takes decisions based on the future predictions over a finite prediction horizon. At the next time step new decisions are taken starting from the new state and over a shifted horizon, leading to a receding horizon policy. Depending on connected devices, the system can be operated in various modes aimed at maximum output without straining the resources.

Particle Swarm Optimization based Algorithm

PSO is a method that was developed for simulating social behaviour inspired by patterns of bird flocking or fish schooling. It was adopted as a computational method to find performance optimization solution by collecting a number of possible solutions (particles) and applying them in the search-space using mathematical formulae to iteratively find their local best position and improve as a solution to the problem. Each particle has a velocity and position that is continuously updated using a time dependant PSO algorithm.

A team of researchers from Taiwan used this method with the roulette wheel re-distribution mechanism to optimize a renewable energy – ESS hybrid generation system. Using real time data for weather forecast, solar irradiance, wind speed, the state of charge of battery and historic data for all of these and load power, a PSO algorithm is used to find if they lie within the boundaries of inequality constraints. A penalty mechanism for the battery keeps in check the depth of charge in life cycle of the battery while the unbalanced power can be reallocated to more superior element to restore balance and achieve lowest accumulated cost.

Artificial Intelligence based Energy Management for Microgrid

Have you watched the movie Her? The Hollywood flick is based on a computer program that has the capability to learn like human beings and become intelligent at properly using the vast amount of global data and supercomputing prowess to interact like a person.

That may happen decades later, but Artificial intelligence has evolved from its phase on paper. Using its various techniques, Aymen Chaouachi and a team of three others demonstrated an artificial neural network learns from the real life scenarios of availability of renewable power throughout the day and load demand. A fuzzy logic expert system is used for battery scheduling and to handle uncertainties regarding fuzzy operation of a microgrid. The approach could be used to considerably bring down operational costs, seamlessly integrate renewable and battery energy sources into the grid.

The Harmony Search Algorithm

Metaphorically, each musician plays a note in way to create most harmony among themselves. In metaheuristic terms, Harmony Search is an algorithm that uses a random initial population (HM) and objectively improves it against three key ideas – memory consideration, pitch adjustment and random research, until the termination criterion is reached. Nikam T. and KAvousi-Fard A. used it to create a suitable optimizing framework for a microgrid system with renewable energy sources and battery.

The algorithm is initialized with data sets like load pattern, price pattern, output power capacity, renewable energy source’s forecast power, magnitude of voltage, bus data and network topology. Initial population or the HM matrix is created next. The best performing parameter is found and objective functions for each solution to the parameters is combined into a single value using fuzzy min-max approach. If new solution is better than the worst vector, the parameter is updated and the algorithm is re-run.

The report stated that IEEE test system showed superior performance with optimal scheduling and reconfiguration of the power units in the microgrid. This lead to an improved overall system performance from the view of all objectives while needing drastically low computations power.