UPTEC ES09013
Examensarbete 20 p
September 2009
Pre study of lead acid battery
charging for wind power
Teknisk- naturvetenskaplig fakultet UTH-enheten Besöksadress: Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0 Postadress: Box 536 751 21 Uppsala Telefon: 018 – 471 30 03 Telefax: 018 – 471 30 00 Hemsida: http://www.teknat.uu.se/student
Abstract
Pre study of lead acid battery charging for wind power
Magnus Vidmo
This thesis is a pre-study of lead acid battery charging for variable speed generators connected to vertical axis wind turbines. A system that controls the turbine to optimize the energy absorption while the batteries are charged at a healthy and efficient way is proposed.
The system is made for applications that are sited far away from the main grid, such as vacation cottages, boats, caravans and radio base stations. The system should be able to work without maintenance for periods up to a year.
The thesis includes theory of aerodynamics, lead acid batteries and battery charging. The main subjects are the optimization of the energy absorption from the wind, how to obtain a long battery life and the integration of a battery bank in the system without interfering with the consuming load. The system is going to be built and tested with a vertical axis wind turbine in Marsta north of Uppsala.
Acknowledgements
Thanks to
Lars Högberg for his enormous patience
Hans Bernhoff, Mikael Bergkvist and Mats Lejion for giving me the opportunity to work with this master thesis
Kristina Edström for all the contributed battery knowledge Erik Dore’ for good company during late evenings
Sammanfattning
Förnyelsebar energi är ett viktigt ämne i den politiska debatten i dagsläget. Den svenska regeringen har satt ett mål att bygga ut vindkraften till en produktion på 30 TWh/år innan 2020. På Ångströmslaboratoriet i Uppsala utvecklas en ny sorts vertikalaxlade vindkraftverk.
Den största skillnaden mellan vertikalaxlade och horisontalaxlade vindkraftverk är att vertikalaxlade inte behöver vridas efter vinden, de är tystare och den tunga generatorn kan sättas på marken istället för längst upp vid turbinen. En del av forskningen för de vertikalaxlade vindkraftverken görs för kraftverk som inte ska vara nätanslutna. Energin måste då lagras på något sätt. Detta examensarbete är en del av den forskningen.
Energilagring i batterier är väl beprövat, speciellt blybatterier tack vare den stora bilindustrin. Det finns en mängd olika typer av blybatterier gjorda för olika ändamål. Dessutom finns det många olika sätt att ladda ett batteri på.
Syftet med examensarbetet är att utveckla ett system som klarar av att kontrollera ett vindkraftverk samtidigt som en batteribank laddas på ett optimalt sätt. Vindkraftverket ska kontrolleras genom att batteriet laddas mer eller mindre. På detta sätt kan man hålla en förutbestämd rotationshastighet på turbinen optimerat för att utvinna så mycket energi ur vinden som möjligt. Förutom detta har en för ändamålet passande batterityp valts ut.
Hur mycket energi som kan absorberas beror på hur snabbt turbinen snurrar relativt vindhastigheten. För att styra turbinen så att den ligger på en hastighet som alltid motsvarar optimal energiabsorption tar man ut mer eller mindre effekt ur generatorn. Turbinens hastighet beror av dess rotationsenergi, om det dras mer effekt ur
generatorn så kommer turbinen att snurra långsammare och vice versa.
Själva styrningen av vindkraftverket sker i en så kallad mikrocontroller, vilken kan liknas vid en primitiv dator. Mikrocontrollern är programmerad med en önskad laddningsalgoritm som talar om hur batterierna ska laddas beroende på hur
uppladdade de är för tillfället. En laddningsalgoritm är en slags karta vilken batteriet ska laddas efter under hela uppladdningsförloppet. För att laddningen ska kunna styra hastigheten på turbinen visar laddningsalgoritmen endast en maximal ström eller spänning som batterierna kan laddas med för tillfället. Det kan alltså, och kommer oftast att laddas med en styrka som ligger under den maximalt tillåtna. Det betyder att batterierna kommer att laddas under längre tid än vad de skulle ha gjort ifrån en kontinuerlig energikälla.
För att ett batteri ska laddas optimalt och få lång livslängd och hög kapacitet ska det inte laddas med för höga strömmar. Det blir därför viktigt att dimensionera
mikrocontrollern. Det är alltså där som själva effektstyrningen sker vilket leder till att turbinen hålls vid en önskad hastighet.
För att förtydliga laddningssättet så kan det sägas att batterierna laddas med den effekt som finns tillgänglig för tillfället beroende av den rådande vindhastigheten men de laddas aldrig över ett visst tak som är satt av laddningsalgoritmen.
Om mer energi måste dras ur generatorn för att hålla turbinen i rätt hastighet än vad laddningsalgoritmen är satt till vid ett visst tillfälle så måste energin dumpas. Dumpningen av energi kan t.ex. göras med värmeelement eller kylfläktar.
I kontrollsystemet som innefattar elektronik samt batteribanken är batteribanken en stor kostnad. Eftersom investeringskostnaden för batterierna blir en mycket stor del av den totala kostnaden så är det viktigt att batterierna drivs på ett sätt som ger en lång livslängd. Parametrar som påverkar ett batteris livslängd är temperatur,
laddningshastighet, urladdningshastighet, antal genomgångna cykler samt hur djupa cyklerna är. En cykel är ett laddnings och urladdningsförlopp. Djupet på en cykel beskriver hur mycket energi som tagits ur batteriet i förhållande till den maximala batterikapaciteten. Det är alltså många parametrar som påverkar ett batteris livslängd, dessutom påverkar de flesta parametrarna varandra. Allt detta måste tas hänsyn till när laddningsalgoritmen skapas.
Det finns många olika batterityper. I detta arbete är endast blybatteriet omnämnt. Blybatteriet är ett mycket robust och prispressat batteri med en lång historia bakom sig. Den långa bakgrunden och det förhållandevis låga priset gör blybatteriet till ett självklart val i ett samanhang som detta. Det forskas mycket på andra batterityper som har bättre egenskaper för t.ex. livslängden, men än så länge är dessa alldeles för dyra. För många användningsområden är det viktigt att det krävs så lite underhåll som möjligt. Batterier som inte kräver mer underhåll än cirka en tillsyn per år är önskvärt. En batterityp har tagits fram under de senaste åren som gör att kontinuerlig
vattenpåfyllning inte längre behövs. Dessa blybatterier kallas för VRLA batterier (ventilreglerade blysyra batterier). De är slutna batterier med en innesluten process som gör att det vatten som förgasas återförvandlas till vatten igen och därför behålls i batteriet. Dessa batterityper blir därför ett lämpligt val för applikationer som ska klara så lite underhåll som möjligt.
Det ställs även vissa krav på batteriernas egenskaper när det gäller cykling och att tåla djupa urladdningar. För dessa ändamål är truckbatterier väl anpassade. Denna typ av blybatterier klarar både djupa och många cykler innan de måste bytas ut. De är byggda för att cyklas en gång om dagen och för att kunna laddas snabbt till
Abbreviations
VAWT Vertical Axis Wind Turbine
HAWT Horizontal Axis Wind Turbine AC Alternating Current
DC Direct Current
OCV Open Circuit Voltage TSR Tip Speed Ratio
IGBT Insulated Gate Bipolar Transistor
MOSFET Metal Oxide Semiconductor Field Effect Transistor PWM Pulse Width Modulation
SOC State Of Charge
Table of contents
1. Background ... 6
1.1. Research projects for Wind turbines at Uppsala University that will be involved in this thesis... 7
1.2. Earlier thesis for the project ... 8
1.3. Purpose ... 10
1.4. Method ... 10
1.5. Problem formulation ... 10
1.6. The thesis structure ... 10
2. Theory ... 12
2.1. Aerodynamic theory for wind power ... 12
2.2. Load control of a VAWT ... 15
2.3. Consequences of variable speed VAWTs for stand alone consumers ... 17
2.4. Battery theory ... 18
2.4.1. The chemical reaction in the lead acid cell ... 18
2.4.2. The construction of the lead acid battery ... 19
2.4.3. Important parameters for the battery operation ... 21
2.4.4. Charging and discharging for lead acid batteries ... 30
2.4.5. Cycling and lifespan of the battery ... 38
2.4.6. Different available battery types ... 39
2.4.7. Connections of batteries for a battery bank ... 41
3. Result ... 42
3.1. Summary of the battery theory ... 42
3.2. The system configuration ... 43
3.3. Choice of charging algorithm for the final solution ... 46
3.4. Choice of battery type ... 49
3.5. Battery bank dimensioning ... 50
1. Background
Nowadays renewable energy is growing strongly on the energy market. In Sweden the government has a development goal to install wind power that gives 30TWh/year before 2020. [1] At the Ångström laboratory, Uppsala University, new strategies for wind power are being developed. The red thread in the research is to use simple and robust constructions to optimize the total performance of the entire system. That means usages of minimal amount of different parts and minimal amount of mechanical parts to reduce maintenance. The main ideas to achieve simple and reliable generators are to use directly driven permanent magnet generators with variable speed, that’s a slow rotating generator with many poles. This technique doesn’t include a gearbox. At Uppsala University the main line for wind energy converters are vertical axis wind turbines VAWT, see Figure 1. The big difference between a VAWT and a Horizontal axis wind turbine HAWT is that the VAWT doesn’t need to be adjusted after the direction of the wind and the heavy generator can be put on the ground. Besides this the VAWT has a lower noise level than the more conventional HAWT.
Figure 1. Vertical axis wind turbine [22]
This thesis is a part of a project for wind power where grid connection isn’t possible. Instead of using a diesel aggregate for power source, the VAWT will be used together with a large battery bank where the received energy will be stored.
There is a large difference in the shape of the entire system if the system is grid connected or not. The biggest difference is that the energy must be stored somehow. The storages are also a very expensive part of the system and it is therefore very important to maximize the batteries capacity.
1.1.
Research projects for Wind turbines at Uppsala
University that will be involved in this thesis
In 2006 a VAWT was built in Marsta north of Uppsala. Marsta has been a place for meteorological studies for many years, and now it also serves as a research place for VAWTs.
The first VAWT that was built at Uppsala University is called Lucia, see Figure 2. Lucia is six meter high with three 5 meter long fixed blades. It generates 12 kW at a wind speed of 12 m/s. The generator is a permanent magnet generator placed on the ground with the axis connected to the turbine. The prototype is available for different tests and research projects in the search for the best working technique for an entire wind power station. [2]
Figure 2. The Lucia turbine in Marsta. [2]
In 2008 another research turbine was built in Marsta named Birgit see Figure 3. It’s a 10 kW VAWT with a four-bladed H-rotor, made for Ericsson’s new radio
Figure 3. The Tower tube turbine in Marsta. 1
1.2.
Earlier thesis for the project
In an earlier thesis for the same project a control system circuit was made by Lars Högberg. This thesis is among others based on the former work done in [4]. The earlier circuit below (Figure 4) has been reformed to operate in desired way for the battery charging system.
1.3. Purpose
The purpose of this master thesis is to develop a cost-effective method to run a battery bank in the best possible way for off grid consumers.
1.4.
Method
To meet the purpose, a study of the earlier built system and reconfiguration of it have been done. To know how the new system should be built in the best way to keep the cost down and to maintain a long life for the batteries, theories for the following subjects have been investigated:
-load control for the VAWT -DC/DC step down converters
-Parts and their functions of the lead acid battery
-Different available lead acid batteries on the market today and which one that will suit this project best
-Different charging methods available for lead acid batteries -Conditions affecting battery life time
Search to find and buy a suitable DC/DC converter and other needed parts for the configuration have been made. This thesis will later result in a built and tested system for the Tower tube wind mill in Marsta north of Uppsala.
Apart of this, a program have been made in Matlab that calculate the storages capacity needed for a special area only using wind data of the specific area. For different sizes of the battery bank the amount of diesel to maintain the energy needed can also be calculated in the program. Results of the program can be seen in appendix 4.
1.5. Problem
formulation
The major issues to solve for the project are:
-How is it possible to control the load for the turbine while it’s connected to a battery bank?
-How will the batteries be affected of the new load control system?
1.6.
The thesis structure
parameters that are important for the battery charging, discharging and the effects of the load control are mentioned. Different types of batteries that are of interest for the purpose are mentioned and also different ways of charging and discharging.
2. Theory
2.1.
Aerodynamic theory for wind power
To extract the most possible power from the wind the airfoils should have a special design. The design is important but the winds angle of attack at the blades is also very important. The angle of attack is the angle between the chord line and the relative wind, see Figure 5. When the blade is moving it will feel a wind that is caused by its own movement. The relative wind is the resultant of the actual wind speed and the wind speed caused by the blades movement. The relative wind will then change when the speed of the blade is changing.
To optimize the extraction of the wind a desired angle of attack should be kept at all time. The wind results in a resultant force on the blades. The resultant force can be divided in a lift force and a drag force, see Figure 5. The tangential force that is in the direction of the blades movement is the force that is contributing to the energy
conversion. The angle of attack that gives the highest tangential force is the desired angle to keep all along.
Figure 5. Angle of attack, Lift force and Drag force for an airfoil.
For a wind turbine there are several different ways to maintain the desired angle of attack. One common way is to have an inbuilt system in the blades that angle the blades in the desired position. This is called pitching. Another way is to keep the turbine at a desired speed which keeps the angle to the relative wind constant. This means that for a specific wind speed there is a specific turbine speed causing the desired angle of attack. In fact there is a typical number relation between the
Cp is a value of the efficiency of the rotor which depends directly on the angle of
attack of the airfoil. To achieve the highest efficiency for the turbine Cp should be as
high as possible. Cp is the ratio between turbine power and the existing power in the
wind. wind the in Power power Rotor Cp (1)
It is impossible for the turbine to extract all power in the wind. A part of the incoming wind will not be absorbed. If all the power in the wind would be extracted by the turbine, all of the incoming air would be accumulated behind the turbine which of course would be impossible (see [5] for a deeper understanding). Because of this there is a maximum value of Cp= 16/27 for a HAWT. This Value is called the Betz limit.
For a turbine that has fixed blades (without a pitching system) the tip speed ratio λ is an important parameter to achieve high efficiency. λ is a number of how much faster the tip of the blade is moving than the undisturbed wind.
speed wind d Undisturbe blade the of speed Tip (2)
Note that for a VAWT the speed is the same for the whole blade.
A specific λ will give a specific Cp for a specific turbine. The optimum value of λ can
be measured and calculated for a specific turbine, see Figure 6.
This means that when the undisturbed wind speed increases, the tip speed of the blade must increase to keep an optimum λ ensuring an optimum Cp.
For the Lucia turbine λ=4 is the optimum tip speed ratio to achieve highest Cp. That
means that the tip speeds of the blades should always be four times faster than the actual wind, but above a given wind speed this is not possible (see Figure 7).
Change of the rotors tip speed
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25
The undisturbed wind [m/s]
T h e r o to r t ip sp eed [ m /s ]
Figure 7. The rotors tip speed changing with the wind speed.
A wind energy converter has limits depending on the construction. The maximum tolerance will be reached at a given wind speed, e.g. the bearings have a maximum tolerance speed. After the point where the maximum tolerance wind speed is achieved the turbine is set to continue to rotate at the same speed. After this point λ will
decrease and Cp will follow causing a lower efficiency. In Figure 6 the Cp value will
slide down to the left of the top.
The power in the wind that the turbine can absorb is
3 2 1 v AC P p (3)
where δ is the density of the air, A is the swept turbine area and v is the wind velocity. [5]
of the wind power station. The diagram also shows when the Lucia turbine will start to extract power. This will be done when the wind reaches 4 m/s. [6]
Figure 8. The theoretical power outtake of the Lucia turbine in kW. The red dotted line shows the power absorbed if there were no construction limits of the wind power station. Between 4-10 m/s the tip speed ratio (TSR) is 4. After 10 m/s the turbine speed is fixed. [6]
2.2.
Load control of a VAWT
As said above the turbine should be kept at a specific speed specified for a certain wind speed to achieve a high Cp and also not to destroy the turbine. The system that
controls this is called the load control.
The tip speed of the blade is measured adjuvant with the voltage from the generator. The voltage is proportional to the rotor speed,
dt d N
where U is the generated voltage, N is the number of turns of the windings, and
dt d
is the time derivate of the alternating flux.
The power for a rotating body is dependent of the torque and the angular velocity
P=M*ω (5)
where P is the power, M the torque and ω the angular velocity. The power outtake is the voltage multiplied with the current
P=U*I (6)
where I is the current. Equation 4 and equation 5 gives that as said before U is proportional to ω and I is proportional to M.
The control system is made to take out more or less power from the turbine to control the turbine speed. If the turbine is rotating too fast, above the desired speed, a higher power outtake must be made to slow down the turbine. When the turbine is rotating to slow less power is taken out, and the absorbed wind power will be stored in the rotating turbine causing it to speed up.
The control is made to keep the turbine at fixed rotation speed during short time periods. During longer periods of several minutes the controlling mechanism is set to keep the turbine at a mean value of λ.
To understand the controlling mechanism better the function “pulse width modulation” (PWM) must be understood.
PWM is a tool that gives the opportunity to control a parameter to approach a desired value. For example a DC/DC converter gets an in value of Uin=100 Volts and the
output Uout should be 48 Volts. The desired value is then 48 V. A control value the
duty cycle D is then multiplied with Uin to get Uout, as in equation 6. The duty cycle
yields a signal to a switching device. D is a value of how much of the time the switch should be open or closed.
Uout=D*Uin (7)
D is set with a control signal set by typically a triangle wave. Uin will pass the triangle
wave in some height. Depending on where Uin passes the triangle wave D will get a
2.3. Consequences
of
variable speed VAWTs for
stand alone consumers
Intermediate energy sources like wind power have one major issue. They produce a fluctuating power. At the main grid this is solved by regulating power as e.g. hydro power. However, this will be a problem for island-operated systems. The load demand will not be the same as the power from the VAWT.
To solve this problem the excess energy has to be stored one way or another, or be dumped as heat. When the power is too low, some other energy source has to support the load. The support and storage device could e.g. be a battery.
A load has regulation for its input voltage. A variable speed machine gives different voltages depending on the present wind speed. These types of machines are supported by power electronics to give a desired voltage and frequency when connected to the main grid. This is also the case when the machines work for stand alone consumers, see Figure 9.
The fluctuating voltage from the VAWT is controlled by a DC/DC converter that supplies the load with the desired voltage. If the power from the VAWT is too low, the external source must be able to supply the desired power for the load. A battery gives a higher voltage when fully charged compared to when partly discharged. The battery voltage range must be in the desired range of the load.
The wind is constantly changing. The optimum would be to regulate the turbine speed according to the exact wind speed at all time. This is not possible in reality. The control system that regulates the power outtake has to have some kind of time step. It’s neither possible nor necessary to regulate the turbine after all the rapidly changing gusts. It’s very difficult to make accurate measurements of the wind speed during short time steps. The turbine also has a large moment of inertia which reduces the possibility of fast regulations.
A mean wind over a longer period is used which will keep the frequency of the control system low. The result of this is that λ will not be at optimum during strong short gusts and Cp will be lower, but the turbine will run smother. Frequencies that are
harmful for the load should be avoided or filtered.
2.4. Battery
theory
Battery theory in this thesis only concerns the lead acid battery.
2.4.1. The chemical reaction in the lead acid cell The lead acid battery has lead dioxide (PbO2)and metallic lead (Pb) as active material
on the positive and negative electrode, see Figure10. The electrolyte is a sulfuric acid (H2SO4) solution of typically 37 weight percent of acid. During the discharging state
the lead and lead dioxide are consumed with the sulfuric acid solution to produce water and lead sulphate (PbSO4). The lead sulphate will be formed at the electrode
surfaces. When the battery is charged the process reverses, the lead sulphate is broken down resulting in lead dioxide and lead recovered to the positive electrode respective to the negative electrode. The chemical reaction for discharging is as follows:
For the negative anode: Pb Pb2+ + 2e
Pb2+ + SO42- PbSO4
For the positive cathode: PbO2 + 4H+ + 2e Pb2+ + 2H2O
Pb2+ + SO42- PbSO4
Figure10. Schematic picture of the lead acid battery being discharged.
2.4.2. The construction of the lead acid battery
The battery’s equivalent model is shown in Figure 11. The inductive reactance of the battery is neglected for low frequencies, while the capacitive reactance is increasing with lower frequencies. More about the battery parameters will be explained later. [8] [9]
Figure 11. Model of a battery according to Randles’ battery model. C1 is considered to be the
main charge storage, R3 is the self discharge resistance, R1 is the resistance of the battery’s
terminals and inter-cell connections and C2 and R2 are results of shifting electrolyte
concentrations and plate current densities. [8]
The capacity of the cell is proportional to the surfaces area of the plates that are in contact with the electrolyte. The plates are often made as a paste with a lot of pores that increases the active plate area. The active material is fixed by a grid which works as a mechanical support and current conductor. This construction is called the flat pasted plate and is very common for the negative plate. The positive plate is
commonly either a flat plate or a tubular plate. The tubular plate has many advantages for heavy cycled operations. The paste are there held in micro porous tubes that are connected in series. To increase the capacity even further plates can be parallel connected to achieve a larger active area.
The grids are made of lead alloys. The lead is not strong enough as a support for the active material. Metals as antimony, calcium, selenium or tin can be alloyed to improve the grids. The alloys change the properties of the plates in different ways.
The antimony alloys can be deep cycled more times than the calcium alloys. Tin added to lead-calcium alloys improves the cycling capability.
The calcium alloys have lower self discharge rate.
If the cells are long time overcharged the positive calcium alloyed plates will grow due to oxidation which can cause the cell to be damaged
Adding Selenium gives plate properties between the antimony and the calcium alloys.
The most used alloys are antimony and calcium.
The separators main function is to electrically insulate the plates from each other. The electrolyte which makes the electron transport possible has to be able to pass through the separator. The separators construction is therefore very important. [10] [11]
2.4.3. Important parameters for the battery operation Parameters mentioned here are different specified voltages, battery losses, internal resistance, battery capacity, C-rate, SOC, the electrolyte density and the temperature.
Voltage
The theoretical voltage of a battery is a function of its construction and the
surrounding temperature. The cathode and anode materials and the structure of the electrolyte are parameters that all matters for the voltage value.
The open circuit voltage OCV is a function of the electrolyte concentration and the temperature. The OCV is a close approximation of the theoretical voltage which is 2.125 V per cell for a fully charged cell with a 1.28 kg/dm3 electrolyte concentration. The OCV is the voltage that is given when the current is zero.
The cut off voltage where the battery is said to be fully discharged is around 1.75 V/cell. The cut off voltage is specified by the manufacturer and is a function of the discharge rate and the temperature.
During charge the charge voltage is between 2.3 and 2.8 V/cell depending of the charge algorithm.
The gassing Voltage is the voltage when the temperature is so high so the electrolyte will start to gas. The cell gassing voltage is specified in Figure 13 for different temperatures.
Figure 13. Cell gassing voltage at different temperatures. [10]
Losses of the battery cell and the internal resistance
Losses from the battery are dependent of the polarization and the internal impedance. The losses are given as waste heat. A lead acid battery has a typical efficiency of 85% from charging to discharging.
There are two different polarization phenomena. The activation polarization drives the electrochemical reactions at the electrode surfaces. The concentration polarization is due to the concentration difference of the products and reactants at the electrode surface.
The internal impedance also consumes energy and reforms it to waste heat. The internal impedance causes a voltage drop which is proportional to the current drawn from the system. The battery’s total internal resistance which follows ohm’s law is a sum of the characteristics of the electrolyte, the active mass, the plates, the straps, the terminals and the contacts between them. The internal resistance is an indication of how deteriorated a battery is. The resistance will rise when the battery is aging. The internal impedance rises during discharge and decreases during charge due to the electrolyte’s changing chemistry.
The losses due to polarization and internal impedance rise with higher currents, which can be seen in Figure 14. The power losses are dependent of the current in quadrate.
on polarizati
R I
P 2 (8)
To get most possible theoretic capacity out of the battery a low current should be drawn from the battery. Although there is a limit for how low the current should be. If the current is as low as the polarization current then the losses will remarkable big.
A comparison to the conventional car engine can be made. An engine that will be driven at low speed will consume less gasoline than an engine that will be driven at high speed, covering the same distance.
Battery capacity, C-rates and state of charge (SOC)
The battery capacity can be described in many different ways. Therefore is it important to specify exactly what kind of capacity that is being discussed to avoid misunderstandings. The maximum capacity available in a battery is determined by the battery’s quantity of active material, the surface area of the plates and the amount of electrolyte. These factors only depend on how the battery is constructed, but the battery capacity also depends on how the battery is used.
The battery capacity is often rated after the current rate. The capacity is different depending of the current rate during discharge and charge, see Figure 15.
Figure 15. The discharge rate effect on the capacity for traction batteries with tubular plates and flat pasted plates. [10]
The rate is called the C-rate. A battery that is discharged with a rate C1 is totally
discharged with a constant current during one hour. If it’s a C5 rate the battery will be
discharged with constant current during 5 hour’s. The C5 rate has a lower current rate
but the C5 rate will also deliver more energy than the C1 rate out of the same battery.
The nomenclature of C1 is equal to 0.1C or C/1. The charging current rate is
mentioned in the same way. A C20 battery is charged during 20 hour’s.
An extension of the nomenclature is e.g. 0.1C5, which means that a battery that is
It’s important to notice that the capacity is specified for a specific C-rate, while retailers have different standard rates when they specify the battery capacity. A Sonnenschein 602 battery can e.g. deliver 12 A during 100 hours (C100 rate) giving a
totally capacity of 1200 Ah at 20 °C. If the same battery delivers a constant current of 100 A it can only be discharged for 10 hours (C10 rate) before it reaches the cut of
voltage, the total capacity is then only 1000 Ah. Note that the battery that is rated to 1000 Ah during a constant current can still give more energy if the current will be lowered, see Figure 17. Even though, it can never give the same amount of energy as the one which was discharged with a lower C-rate. There will always be larger losses for the higher C-rates. [10] [11] [12]
When a battery is totally discharged it has reached its cut of voltage. After this point the voltage will continue to fall but there is not much more energy left to drain and the battery will only be damaged. The cut of voltage is different dependent off the C-rate and the type of battery. Higher C-rates has a lower cut off voltage as can be seen in Figure 16.
Figure 16. The discharge time in relation to the discharge current. The discharge time is
measured until the cut off voltage is reached. The cut off voltage is following the dotted line in the figure and is dependent of the discharge current. The figure is only an example of a specific battery and will differ for different batteries.
Figure 17. The voltage for a battery discharged at sequentially lower C-rates.
The battery capacity left to use at a given time for a specific C-rate is often stated as the state of charge SOC. The SOC is a percent value where the Ah discharged and charged from the battery is divided with the nominal capacity of the battery at a specific C-rate. If the battery will be discontinued discharged with moments of charging the SOC formula will be
NomCap dt I dt I NomCap SOC ch e e ch e disch e disch arg arg arg arg 1 1
(9)where NomCap is the nominal capacity of the battery at a specific C-rate, Idischarge is
the discharge current, Icharge is the charge current, ηdischarge is efficiency of the
discharge and ηcharge is the efficiency of the charge. The overall efficiency of the
battery charged and discharged is a quota between the energy delivered during charge and the energy received during discharge. It’s difficult to measure ηdischarge andηcharge
individually.
The usage of SOC can be a bit confusing while the nominal battery capacity is stated at a specific C-rate. The C-rate that the SOC is stated for is typically the C-rate close to the optimum C-rate where the battery can deliver a maximum possible amount of Ah. This means a current rate so low that there will be minimum possible losses. If a battery is rated to 200 Ah at a specific C-rate and it’s at 80% SOC then it can still deliver 160 Ah before it reaches the cut of voltage at that specific rate.
SOC and the aging batteries SOC for SOCage. The different SOCs are good parameters for different things for e.g. the charge algorithm. [10] [13]
The SOC is either measured with the assistance of the electrolyte density, the OCV or with the equation 9. Note that the density measurement can not be done with VRLA batteries. One way to know when the battery will be fully charged is to measure when the current decreases till its gets constantly small over a longer period.
The electrolyte density and the surrounding temperature influence
It’s important for the life span of the battery to know what effect the reaction between the different substances in the battery construction will have. The sulfur acid solution in the electrolyte is aggressive to some separators and some other components in the battery. Higher concentration of the sulphur makes the electrolyte more aggressive. Lower concentration makes the solution less conductive. Therefore a density value that isn’t to aggressive but still gives a good conductivity is preferable.
The solution also makes the lead corrode. The amount of corrosion is due to the sulphur concentration. In temperate climates the electrolyte solution has a weight percent of 37 % and a density of about 1.26-1.28 kg/dm3. A matter of fact is that an electrolyte density of 1.28 is least corrosive for the lead. Therefore to enlarge the total lifespan it is very important to avoid having a battery discharged longer than
necessary.
The electrolyte density is designed after the sector of application. When the battery is fully charged, the electrolyte density is set after the conductivity needed to get a desired capacity out of the battery. The heavily cycled batteries have the highest density and the stationary that are low cycled have less density. For stationary
batteries with small high rate demands and larger proportional electrolyte volumes the concentration can be held lower and be less aggressive. These batteries can have a concentration as low as 1.21 kg/dm3.
The density is also important considering the freezing point of the battery. The freezing point will be lower for higher sulphur concentrations up to a density of 1.3 kg/dm3 where the lowest freezing point at -70 C° is reached. A battery that is fully charged at a density of 1.28 kg/dm3 has a freezing point around -65 degrees and when
that type is fully discharged at a density of 1.16 kg/dm3 the freezing point is around -17 C°. A battery that is fully charged at a density of 1.21 kg/dm3 will freeze at around -27 C° and that type of battery can fully discharged have a freezing point at almost 0 C°. Therefore the electrolyte density is held a little bit higher in colder areas than in warmer areas. [10]
Higher density also results in a higher boiling point.
measured at 20 or 25 ºC. Note that the higher C-rates are more affected by the temperature.
Figure 18. Available capacity in relation to the surrounding temperature at different C-rates. The values are measured at a Sonnenschein A600 battery. [14]
Low temperatures are good for batteries that stand unused for long times. The self discharge will be low when the chemical reduction is held low. A high temperature is therefore not always a good thing, see Figure 19. Thermodynamically the discharged state is most stable and the self discharge will be a problem at all time when the battery is unloaded or not at charge.
Although the capacity is increasing with higher temperatures the total age of the battery will be less with higher temperatures. The battery will hold for fewer cycles with higher temperatures. Batteries that are constantly used at very high temperatures should be built with another electrolyte density for longer life. [10] [11]
The density of the electrolyte is also dependent of the surrounding temperature. The density increases with lower temperatures as the electrolyte contracts by the colder temperature. The density can be calculated with the equation
) 15 ( 10 ) 15 ( ) ( 5 t C t (10)
where δ is the density, t is the temperature in ºC where the density is calculated at and α is a temperature coefficient.
The internal resistance of the battery is dependent of the density of the electrolyte. The internal resistance is both dependent of the temperature directly as said earlier and also as a function of the density that change with the temperature. The density fluctuation due to a temperature shift is however relatively small, (see equation 10) and the internal resistance dependency will therefore be much smaller than the direct dependency of a temperature shift.
The open circuit voltage is dependent of the electrolyte density and the surrounding temperature, see Figure 20 for the density dependence at 25ºC. It’s almost a linear relation between them down to a concentration of about 1.10 kg/dm3. For an accurate measurement the battery should stand alone for 4-8 hours before the measurements can be done. The approximation equation for the linear relation is
OCV = δ (25°C) + B (11)
where B is a constant. B=0,845V. The temperature dependence is about +0.2 mV/ºC for most batteries. [10] With the temperature dependence the equation will be
Figure 20. The cells open circuit voltage as a function of the electrolyte concentration at 25ºC.
The AGM batteries have slightly higher OCV than the curve in Figure 20 shows. Calcium doped batteries often VRLA batteries have 5-8 % higher OCV. [9] [10] The density of the electrolyte is an indication of the state of the battery, see Figure 22. When the battery is discharged the electrolyte density decreases in proportion to the amount of Ah that is discharged from the battery. The state of charge for different types of lead acid cells can be seen in Figure 21. While there is a linear relation between the OCV and the electrolyte density from 1.10 kg/dm3 the OCV can also be an indicator of the SOC.
Figure 21. Electrolyte density at different SOC, for different types of lead acid batteries. More about the different types of batteries can be read about in 2.3.6.
Figure 22. The electrolyte and SOC dependence during constant current discharge and charge.
From the chapters above a conclusion can be made that all the parameters are dependent of each other. It’s good to know all this parameters dependency while measure any of them. See Figure 31 in the result chapter for an overview of all the parameters impact on each other.
2.4.4. Charging and discharging for lead acid batteries
Charging
There are some easy but important rules for battery charging.
If the battery has been deeply discharged it should be carefully charged at the beginning with a proportionally low current.
In the end when the battery is charged until 100% SOCage the battery should also be charged with a low current, normally below the C20 rate.
The most important rule when charging is to avoid the gassing voltage. Many
chargers include rectifying equipment that causes an AC ripple with the direct current. The ripple causes further heating of the battery. It’s important to minimize the ripple especially in the end of the charge where the margin to the gas voltage is less. A way to minimize the ripple is to implement a filter of suitable size. The current that theoretical can be applied to the battery without reaching the gas voltage is an inversed exponential function
where Ahdischarged is the ampere hours discharged at any chosen time, t is the time and
e is the exponential function. This means that the charge current can be the same value as the ampere hours discharged from the battery, e.g. a 100Ah battery that have been discharged with 80Ah can be charged with 80 A. This is the theoretical values. In reality there are other factors that decrease the charge current. To high currents will change the morphology of the electrode and the conducting material will be heated. Side reactions will occur for batteries that will get to warm particularly above 55 °C e.g. the corrosion rate will increase. The losses would be really high, see Figure 14. In reality the charger often have a current limit in the beginning of the chare algorithm. It’s common that the chargers give a constant current up to 80 % of the SOCage. [10] [11]
The current that a battery will be charged at is a proportion of the difference of the battery’s open circuit voltage and the charge voltage, see Figure 23.
ernal OCV e ch e ch R U U I int arg arg (14)
Where Ucharge is the charge voltage and UOCV is the open circuit voltage. The internal
resistance will fall during charge as the open circuit voltage is rising. The charge voltage is e.g. set by the power given from a generator.
Figure 23. Charging scheme. Ucharge = Uload = Ubattery = UOCV + Icharge ● Rinternal = Icustomer ● Rload.
Constant voltage charging
The old fashion conventional method of charging lead acid batteries, often for car batteries has been to put a constant voltage over the terminals. The voltage is then safely set a bit under the gassing voltage. The current in the beginning will then be very high, see equation 13. However there is a current limit set by the electronics in the charger or by the thermal characteristics of the battery. The charge current will then fall as the open circuit voltage rises. When the open circuit voltage starts to approach the charge voltage the charge current will get remarkable low and it will be very time consuming to charge the remaining capacity to reach 100% SOCage. The last percentage of the SOC is important for the battery’s service time. These old time chargers are therefore not suitable for charging a battery bank to achieve high
standard and a good efficiency. Although if the charge voltage is set just under the gas voltage the battery will reach 100 % SOCage faster. Often the chargers stop at a finish rate with a current at around C20. [15] [16]
Figure 24. Constant voltage charging to the left in the figure. When the charge voltage is kept constant while the OCV is rising the current will taper. Constant current charging to the right in the figure. To keep a constant current the charge voltage must rise simounsly with the OCV rise.
Float charging
Float charging is a type of constant voltage charging held at a low potential. The Float charge state is often done when the batteries are fully charged. The low potential is held just enough to cover for the self discharge.
Constant current charging
Another type of charging is the constant current method. The current is held at a fix value until the battery is fully charged. This is not a common method for charging lead acid batteries. To achieve a fast and effective charge algorithm, adjustments of the current rate is needed.
Trickle charging
Trickle charging is a type of constant current charging. The trickle charging is used to maintain a fully charged battery fully charged. It gives a low current around C100 to
Pulse Charging
Pulse charging is a method developed to minimize the charging time. Tests made on this method also show increased battery life. Basically the method consists of high current pulses. The current is much higher than allowed for the gassing voltage. The specialty with the Pulse charging is that the period of the current pulses are
controllable. The time of the pulses are set so the battery won’t be able to heat up. While the battery is being charged and the SOC is getting higher the pulse time has to be shorter. This is a result of the gas development time. The hydrogen and oxygen development has a time constant that is depending of the SOC. If the current pulse is short enough the time is not enough to produce gas. The current will then only be consumed in the charge reaction.
In this particularly way the applied currents can be much higher. Integrating the current over the time pulses shows that the total Ah charged can be applied in much shorter time than for constant voltage or constant current charging, without harming the battery.
Figure 25 shows a common charging scheme where constant current charge first is used where most of the energy is charged to the battery. This is followed by a constant voltage charge where the rest of the energy is charged.
Figure 25. Conventional charge algorithm with a constant current followed by a constant voltage. [16]
Comparing the Pulse method with the commonly used method in Figure 25 shows that the Pulse charge has most benefits during the later part of the charge. The reason for this is that pulses can be very long in the beginning of the charge when the battery is more resistant to high currents. The long pulses make no big difference to the constant current charging. The high current pulses should be used carefully in the early SOC avoiding damages to the battery.
pulses length decreases. When the battery is fully charged the time duration of the OCV to decay is the same as the charge pulses and the battery is kept fully charged at all time. This can be compared with the float charging. Some newer charges have a Pulse charge method in the end of the charge algorithm, to maintain the battery fully charged. This method has shown good benefits for the capacity and battery life in made tests, e.g. in [17].
Many of the up to date chargers have charger algorithms made of many different steps. Keeping the correct pulse periods during the entire charge, a Pulse charge could be made during the entire charge, keeping the algorithm very simple.
Tests have shown that if the pulses also include a small discharge current after each charge current the charge time can be even lower and also giving a higher life time of the battery. The discharge currents equalize the concentration of the active material in the battery. This reaction improves the charge acceptance of the next current pulse. In the test made in [16] two VRLA GEL 28 Ah batteries were used. The discharge current pulse was applied at 8 % of the charge pulse time. The test shows that the energy being charged to the battery is of the same amount as without the discharge current. Note that the charge current is applied at 8% less time than before, see Figure 26.
Figure 26. Relative charging rates with and without the discharge pulses. The curve starts around 50% where the constant current step is finished. This is because the constant current method and the pulse method is slightly similar in the begging of the charge. [16]
Figure 27. The pulse charging with and without discharging pulses and the conventional constant current/voltage charging. [16]
Even though this test shows good features fore the pulse charging, other battery types might show other results. The technique should be used only with certainty that the gas voltage isn’t exceeded. Modifications of the method can also be required when the battery is aging. [10] [16] [18]
Equalization charging
The equalization charging is done if the electrolyte needs to be equalized. Cells that are connected in series will often differ a bit in the electrolyte density, and the density can also differ within a cell from the bottom and up. Parts that have higher density are heavier and will sink to the bottom. The equalization is achieved by a constant current causing a high voltage around 2.65 V/cell. The voltage is then clearly above the gassing voltage and the electrolyte will start to bubble and it will be mixed around. This operation is clearly not good for all types of batteries while some water will be lost in the electrolysis and a small part will also evaporate. For open flooded batteries that can be refilled with water this is however a good thing to do occasionally. Mixed charge algorithm
Figure 28. Different charge steps for charge algorithms. The red line is the charge voltage and the blue line is the charge current. [15]
Overcharge
The battery is often charged to 110-120 % SOCage to compensate for the losses from the last discharge. It’s important to do this overcharging to remain the capacity of the battery, but it should be done carefully without gassing. Too much overcharge will cause the pressure to exceed above the designed venting pressure. The loss of water and internal heating accelerate the positive grid corrosion. Overcharge have a benefit, it results in an equalization of the electrolyte. [10][19]
Discharging
The discharge of the battery will take place when the charge voltage is less than the battery voltage, see Figure 29.
Figure 29. A circuit scheme over the battery cell and the connected load during discharge. Rload is
the resistance of the load. I is the current drawn from the battery. Rinternal is the internal
impedance of the battery. Uocv is the open circuit voltage. Ubattery is the battery voltage during
load. Udrop is the voltage drop over the internal impedance during load. The voltage drop caused
by the polarization is not showing in the figure, its instead included in the internal resistance. The diode in the figure is preventing back currents.
Figure 29 shows the circuit scheme of the battery cell and the connected load during discharge. The equation for the battery voltage is
Ubattery = UOCV – Upol – Icustomer • Rinternal = Icustomer • Rload (15)
where Uocv is the open circuit voltage and Upol is the voltage drop for the polarization.
Figure 30. Effects of discharge pauses.
2.4.5. Cycling and lifespan of the battery
After each discharge and charge the battery chemistry will change slightly. The battery is ageing and the capacity is slowly decreasing. The reaction is not reversible. There are some parameters that are clearly changing the life of the battery.
C-rate. High C-rate gives shorter life.
Temperature. High temperature shortens the battery life.
Depth of discharge. To deep discharges change the battery chemistry and shortens the battery life
Significant overcharge Number of cycles
It’s important to charge the battery all the way up to 100 % SOCage occasionally
Discharges that are deeper than what the battery is made for cause’s sulphation and grid corrosion at the plates. Some of the sulphate will crystallize on the electrodes reducing the active area which reduces the battery capacity. To dissolve the crystals and to recover some of the battery capacity a high voltage must be applied. More crystals require higher voltage. One way to recover the battery without causing
Significant overcharges also cause grid corrosion and if the pressure gets to high it will cause losses of the electrolyte. [17] [19]
The battery ageing can be measured with the resistance. When the resistance of the battery has increased with 25% from the point when the battery was new, the capacity is reduced from 100% to 80%. The resistance can differ with 8% between VRLA batteries of the same batch. A battery is almost useless when the SOCage at 100% is 80% of the SOC. [9]
2.4.6. Different available battery types
The Lead-Acid batteries are still by far the cheapest and most robust battery on the market providing good performance and life characteristics. For this thesis the choice of battery type is therefore very easy. Comparison and investigation on different battery products will not be a part of this thesis. The lead-acid battery have a market of 40-45% of the sales value in the world including for example energy storages, emergency power, vehicles, telephone systems, power tools, communication devices and as the power source for mining.
The electrical turnaround efficiency is about 80 % that’s comparing discharge energy out with charge energy in. A single battery can have a size of thousands Ah. These are factors together with the relatively low price that makes them suitable for energy storages use. The lead acid batteries have a typically energy density of 30-40 Wh/kg which makes them very heavy, although this is not a problem for systems that are sited on fixed places.
There are several different types of lead-acid batteries constructed for specific applications. Optimizations for different parameters distinguish the different types of batteries. Some parameters are e.g. energy density, power density, cycle life, float service life and cost. [10]
The SLI battery
The most common battery type is the lead acid batteries made for the car industry. These batteries are often called SLI batteries, which stands for start, lighting and ignition. The batteries are often open, so gasses can freely vanish. The open batteries are very robust and reliable batteries. The drawback of the open batteries is that they need maintenance and have to be refilled with water. They have a vent plug where the gases that are formed can escape. The start battery made for cars have a high power density to be able to give the high current needed to start a car. After the start the battery is float charged when the motor is running. The battery will seldom be deeply discharged and the cycle life is not an important factor.
time when water is reproduced lead sulphate is produced at the anode. This causes problem for batteries when they are ageing. A possible way to solve this problem is to include a pulse current interrupt in the charging algorithm that dissolves the lead sulphate again.
There are two different types of VRLA batteries GEL and AGM. In a GEL battery there are some substances added to the sulphur acid solution to make it more solid. This is made to prevent the solution to spread to the surroundings. These types of batteries are very robust and can handle deep discharging cycles very good.
In an AGM battery the separator are made of glass fiber. This special separator keeps the sulphur steady by capillary forces. The AGM can get higher power from a smaller volume because the separator can be made very thin with a low inner resistance. The drawback with this type of separator is that the solution has to have higher sulphur concentration. To handle this, AGM batteries should be charged by a higher voltage. [10][11][15]
Two other classifications other than the most common SLI batteries are the traction and the stationary batteries. These batteries can be of open vented or of VRLA type.
The stationary battery
Stationary batteries are typical made for standby or emergency backup power. There can be long periods where the batteries stand unused. They are often left at float charge, to be fully charged when needed. The electrolyte is often exceeded to minimize maintenance and to be more resistant to gassing. This makes the batteries capacity limited by the positive plates in comparison with the traction batteries that are limited by the acid in the electrolyte. To make the intervals between watering greater a nonantimonial grid is used. To be more sustainable against grid corrosion growth, the positive plate is scaled so it can grove 10 % before the battery will be unsuitable for use.
The traction battery
The traction battery is used in vehicles that are driven electrically or as hybrids. The main difference to the SLI battery is that the traction batteries have to give suitable power through all day while the SLI batteries are built to give a high current for a short start up time. This means that the traction battery should be able to be deep cycled many times without being depleted. The SLI batteries can only manage around 10 deep cycles before they are in bad condition.
flat pasted or tubular plates. The tubular have a numerous advantages during operation such as less grid corrosion, less self discharge, less polarization losses and longer life, but they have a higher initial cost. [10] [11]
2.4.7. Connections of batteries for a battery bank
3. Result
3.1.
Summary of the battery theory
In Figure 31 all the operation parameters of the battery is connected to each other. The arrows points at the boxes that are influenced by a parameter. In this way the scheme also shows secondary dependencies. The battery capacity increases with higher temperatures as a direct dependency, but it also decreases with higher temperature while the battery life will be reduced with higher temperatures. The amplitude of the dependency is not showing, this can be read about in the theory chapters.
Figure 31. An overview of the parameters impact on each other. The parameter that another parameter is dependent of is always rising in the figure. The + and – symbols indicates if there is a negative or positive dependency. E.g. the internal resistance is decreasing when the
3.2.
The system configuration
The system is built to keep the turbine at an optimal speed as much as possible. The charging of the battery bank should not affect the efficiency of the turbine. The charging of the batteries will not be optimal in regard to the conventional charging time. The wind energy is simply not enough at all times, although the charging algorithm can be the same as for conventional chargers. When the energy from the generator is less than what the charging algorithm is demanding in regard to the current or the voltage amplitude the charging will just take longer time. When the battery is fully or almost fully charged the energy from the generator may be larger than what the battery and the load can receive. The excess energy is then simply dumped as heat.
For the vacation cottages etc. the heat can be used to heat the house or the water supply. For the Tower tube the area inside will be heated which can be good for the batteries during winter to maintain a high capacity from them. In warmer countries the energy can be dumped with cooling fans to maintain a longer battery life.
The entire load and charge control system scheme is drawn in Figure 32.
The AC-Voltage from the generator is first rectified to DC-voltage. For the Tower tube VAWT the AC-Line to line voltage is kept at 200 V at a wind speed of 10 m/s with λ=4, and at wind speeds above 10 m/s. 200 V is the highest line to line voltage at a frequency of 75 Hz. At a wind speed of 4 m/s and λ =4 the line to line voltage is 80 V and the frequency is 30 Hz. The DC-voltage lies between 108 V and 270 V
dependent of the wind speed.
The dump load is placed after the AC/DC step. The dump load is a resistive load which is switched with an IGBT. If there is more power available than the customer and the battery can absorb, energy will be stored in the turbine causing it to speed up. When the turbine rotation speed is above a decided value the IGBT will start to switch and energy will be dumped to keep the turbine at the optimal speed. See appendix 1 for a detail declaration of the PWM for the dump control.
Next in line in the scheme in Figure 32 is the DC/DC converter. The DC/DC
converter is needed to get a desired voltage. The output voltage should have typically values for charging a lead acid battery and the DC output must therefore be
adjustable. The voltage level is set by the energy available from the generator and the charge algorithm.
Figure 32. Overview scheme of the load and charge system. The customer current Ic is constant
for a specific load and will only differ slightly with the charge voltage. The battery charge current Ib will fluctuate and is only dependent of the charge voltage, the internal resistance and
When the charge voltage will be lower than the OCV of the battery the battery will be in the discharge mode. The entire load current will then be taken from the battery. When searching after appropriate DC/DC converters we found it very hard to find converters that were adjustable. It was also hard to find converters above 2 kW. The reason for the low power is that most converters are made for the wall outlets which have a voltage of 230 volts. The fuse is often at 10 Ampere which gives 2.3kW in power. This problem can be solved by serial or parallel connecting the converters. A serial connection is proper if the batteries are charged individually, which would give good individual charge properties. A parallel connection would be good if a unit would crash. If a 10 kW system would include five 2 kW units and one would crash there would still be 8 kW left to use. A system that is properly built for the chosen wind site would then still run at optimum until very high wind speeds would occur. For the case that will be built and tested we have obtained six 2 volt cells to build a 12 volt battery bank. The DC/DC converter chosen for that project is a Cosel PBA
1500F-153. The ripple from the converter should be considered if it is at a reasonable
level. If the ripple is too high a filter is required after the converter. The ripple is too high when it affects the charge voltage with about ±0.005 volts.
To be able to control the DC/DC converter in a way that give desired values for the battery and the turbine speed at all time, a battery charging microcontroller is connected to the DC/DC converter. The quota between the wind speed and the
rotation speed indicates how much power that is available for battery charging and the load. The microcontroller is programmed with a charging algorithm which is steered by measured inputs such as the customer and battery current, the battery voltage and the battery temperature. The charge algorithm sets how much of the available power that is going to be used to charge the battery and how much that will be dumped. The charging microcontroller gives a signal to the DC/DC converter of 0-5 volts. For a 12 volt battery e.g. 2.5 volts means that the DC/DC converter should give a voltage of 15 V to the battery.
When the battery is discharged and can’t deliver the desired load current a diesel aggregate will be used. When the diesel aggregate is used the battery bank can be charged again from the VAWT and the gate to the customer will be open. It will be charged until a level where it’s able to give a suitable load current again for a longer time. This will prevent that the diesel aggregate will turn on and off over short time periods.
The aggregate should also be able to charge the batteries in case of bad wind conditions during a longer time. The batteries life time is shortened much faster if they stand completely discharged. The gate from the diesel aggregate should be controlled by the charging microcontroller. When the battery voltage is constant below a specified voltage during a specified time dependent of the battery type the gate will be closed.
When maintenance is done of the system, the state of the batteries can be measured with a rapid battery tester such as the Spectro CA-12.4
3.3.
Choice of charging algorithm for the final
solution
The charge algorithm chosen in this result is made to extract as much as possible of the wind energy in the turbine optimized for a high Cp. The wind is a fluctuating
source, it’s important to always take as much as possible out of the wind energy. Another way to form the charge algorithm would be to make a load control optimized for the battery charging. The WAVT could then be controlled by an optimum C-rate for the battery charging, instead of a specified λ for optimum Cp.
When the specified C-rate for the optimum battery charging is higher than what the generator can deliver at a specific moment the battery C-rate will simply be lower. This is exactly the same case as for the charge algorithm made to keep an optimum Cp.
When the specified C-rate for batteries is lower than what the generator can give at a specific moment the extra energy must either be dumped or be saved as rotation energy in the turbine. If the extra energy is dumped the VAWT can never give a higher power than the optimum C-rate for the batteries multiplied with the DC voltage e.g. 48 volts multiplied with 50 amperes gives 2400 W.
Energy saved as rotation energy can only be saved for very short moments depending of the moment of inertia of the turbine. For both the Lucia and the Tower tube the turbine would have to be shut off in just a few seconds not to destroy the construction due to the high speed and the Cp would decline very fast see Figure 6. This type of
charge algorithm would give much higher loses and the overall efficiency would be much lower. See appendix 3 for calculations of the saved extra rotation energy. To extract most energy possible, equation 13 should be used, although when the batteries capacity is lower than the nominal power output of the WAVT the battery will be damaged due to high currents. E.g. if the VAWT can give 200 amperes at 48 volts and the battery capacity is only 400 Ah the battery will be damaged very fast. This shows the importance of big battery banks to be able to take care of all the incoming wind energy.
Proposal one for a charge algorithm
The first step of the algorithm will be set by a maximum current level. The level is set by the batteries specifications of highest acceptable current. Normally the rate is about C3. The high current will provide high losses, but it’s important to use all the available
energy in the wind.
