Battery Monitoring Why – What - How DIEBOLD POWER SOLUTIONS LLC

Agenda Introduction Why should I monitor? Increase Battery System Reliability Cost savings Safety What technology should be used A “battery monitor” is not enough Resistance vs. Impedance, why does it matter Electrical Engineer version Pipe Analogy How does ripple noise affect my measurements Case study Product description Configuration examples

Based in Boca Raton Florida Who is Alber? Established 1972 Based in Boca Raton Florida We manufacture battery test equipment based on extensive experience in and knowledge about Battery design and Battery aging characteristics. We are the Battery Test Experts!

Battery Monitors, Test Equipment and Educational Services Condition assessment Capacity assessment Education On-line Public Manual Cellcorder Single Cell Tester SCT-200 Battery Basic Seminar Digital Hydrometer DMA-35n Automatic/Continuous Monitor system BDS and MPM Battcon Conference Test Tools RM-650 & PSC-10 Off-line In-house Contact Resistance Micro-ohm meters BCT-2000 Battery Cell Scanner Battery Basic Seminar MLT-12 Momentary Load Tester CLU Continuous Load Units Product training

Capacity vs. Condition assessment. What’s the difference? Battery Capacity Testing. Off-line discharge test using load bank with constant current capability (ability to adjust resistance as voltage drops) while continuously monitoring each individual battery cell voltage to identify the bad cells. Time consuming (1 to 10 hours plus re-charge time) Requires external load bank and sometimes backup battery Reliable. Only way to truly assess the battery’s absolute capacity

Capacity vs. Condition assessment. What’s the difference? Condition assessment. On-line test measuring the internal resistance of each battery cell. As a battery condition declines, the resistance increases. Increased internal resistance indicates aging or pre-mature failure before the cell cause critical failure to the battery On-line, non-invasive test that can be performed manually (Cellcorder) or automatically (Monitor System) Provides absolute state-of-health assessment parameters on individual cell level Detects failing cells before they cause problems Does not provide an absolute capacity value and chemical problems are more difficult to detect in an early stage.

Why? Why Monitor? Batteries International July 2003 “-Batteries are the heart and Achilles heel of critical infrastructure and have been so for the last 20 Years” EYP CTO/CEO Peter Gross.

Why Monitor? Why? US Department of Energy estimates that total annual losses from Power outages for large industries may be as high as 150 billion dollars.

Why? Why monitor? Increase battery system reliability Cost savings Avoiding costly unplanned down time Reliable test data through consistent data acquisition Cost savings Reduced maintenance man-hours or redirection of maintenance crew to proactive maintenance activities Optimizing battery life Safety Limiting personal hands-on and avoiding human errors Avoiding catastrophic battery failures

Why? Why monitor? Increase battery system reliability Avoiding costly unplanned down time Reliable test data through consistent data acquisition

Avoiding costly down time Why? Uptime (9’s) Downtime / year 99.9% (3) 8 hr, 45 min, 36 sec 99.99% (4) 52 min, 33.6 sec 99.999% (5) 5 min, 15.36 sec 99.9999% (6) 31.5 sec 99.99999% (7) 3.15 sec Source: 7X24 Exchange 2003 Spring Conference

Avoiding costly down time Why? Can you afford not to monitor? Industry Avg. Cost Per Hour* Cellular communication $41,000 Telephone Ticket Sales $72,000 Airline Reservation $90,000 Credit Card Operation $2,580,000 Brokerage Operation $6,480,000 * Does not include intangible losses such as damaged reputation and lost customers.

How much would a power outage cost your company? Avoiding costly down time Why? Can you afford not to monitor? Industry Avg. Cost Per Hour* Cellular communication $41,000 Telephone Ticket Sales $72,000 Airline Reservation $90,000 Credit Card Operation $2,580,000 Brokerage Operation $6,480,000 How much would a power outage cost your company? Source: 7X24 Exchange 2003 Spring Conference

Why Monitor? Why? Because Batteries Fail! Batteries are like humans. Sooner or later the battery will reach end of life. It can happen through… “Normal” aging Positive Grid Corrosion Pre-mature failure Manufacturing defects Battery “abuse” High temperature Excessive charge current Defective chargers Damaged battery jars

It only takes one bad cell to bring a string down Avoiding costly down time Why? 0 V 480V It only takes one bad cell to bring a string down

Fixed installed leads captures test data the same way, every time. Reliable test data Why? Fixed installed leads captures test data the same way, every time.

Reliable test data Why? Pre-mature failures develop faster. Monitoring will detect problems earlier.

Why? Why monitor? Increase battery system reliability Cost savings Avoiding costly unplanned down time Reliable test data through consistent data acquisition Cost savings Reduced maintenance cost or redirection of crew to perform proactive maintenance Optimizing battery life Often double useful battery life

Why? Cost savings Optimize battery life Monitor critical parameters and take appropriate action Internal resistance Replace bad cells before they affect other cells Maintain a balanced resistance level in all redundant strings Temperature Temperature has a direct influence on battery life Uneven temperature over the string cause cells to float differently Voltage Charge voltage should be matched to ambient temperature Float current Improper ripple affects the battery life

Why? Why monitor? Increase battery system reliability Cost savings Avoiding costly unplanned down time Reliable test data through consistent data acquisition Cost savings Reduced maintenance man-hours or redirection of maintenance crew to proactive maintenance activities Optimizing battery life Safety Limiting personal hands-on and avoiding human errors Avoiding catastrophic battery failures

Detect problems before they cause a catastrophic system failure! Safety Why? Detect problems before they cause a catastrophic system failure!

Increase Reliability How? Most monitors will be able to detect a failed battery by... monitoring overall voltage or individual cell voltages during discharge if the discharge is long enough to show a drop in voltage trending increased cell impedance if the charger noise is low, if the system load is constant and if no cells are replaced

Increase Reliability How? Most monitors will be able to detect a failed battery by... monitoring overall voltage or individual cell voltages during discharge if the discharge is long enough to show a drop in voltage trending increased impedance or conductance… if the charger noise is low, if the load is constant and if no cells are replaced But it’s a gamble! The battery may FAIL during the next outage because the chosen technology does not alert you to the problem early enough!

How? Increase Reliability What is needed to detect a failing cell before it cause a problem? We have to detect problems in the conduction path Internal to the cells Inter-cell connections. Internal resistance is directly related to the battery cells capability to generate power

How? Ohmic measurements The following terms are used, in the battery industry, to describe internal ohmic measurements : AC Impedance (Z) AC Conductance (Siemens) DC Resistance These are the three different terms used in connection with Ohmic measurements. Please note that AC Impedance should be simply called Impedance. The definition of Impedance is opposition to AC current flow. Therefore AC Impedance is redundant AC Conductance is not correct; the definition of Conductance is the inverse of Resistance. The commercially available AC Conductance meter actually measures the same thing as Impedance except it inverts the answer. In this presentation it will be treated as an Impedance meter. The presented value and entity is not the issue, it‘s a question about AC vs. DC based measurements

Cell Resistance If a 2 V Cell’s internal resistance increases more than 25% above its baseline value, that Cell will not pass a 3 hour capacity test. It may pass an 8 hour capacity test but it will NOT pass a 30 min capacity test Capacity Most battery manufacturers base their warranty settlements on a 50% increase in the ohmic value and as will be pointed out later, that is not a valid number for maintaining a reliable battery system. Resistance

Internal resistance Why is internal resistance related to battery capacity and why is it important to use an appropriate measurement method? New Grid Aged Grid

Select illustration for Push on the button below to jump to the illustration you prefer Electrical Engineers Non Electrical Engineer

Simplified equivalent circuit The conduction path through a battery includes the: Resistance of the Post, Strap, Grid, Grid-to-Paste, Paste, Electrolyte, Separator and so on... This is the generally accepted simplified circuit equivalent circuit of a lead acid cell

Simplified equivalent circuit The conduction path through a battery includes the: Resistance of the Post, Strap, Grid, Grid-to-Paste, Paste, Electrolyte, Separator and so on... The cell also has a huge capacitor C This is the generally accepted simplified circuit equivalent circuit of a lead acid cell

Simplified equivalent circuit The conduction path through a battery includes the: Resistance of the Post, Strap, Grid, Grid-to-Paste, Paste, Electrolyte, Separator and so on... The cell also has a huge capacitor This capacitor is connected in parallel over about 45% (R2) of the total resistance R1~55% C R2~45% This is the generally accepted simplified circuit equivalent circuit of a lead acid cell

Simplified equivalent circuit The conduction path through a battery includes the: Resistance of the Post, Strap, Grid, Grid-to-Paste, Paste, Electrolyte, Separator and so on... The cell also has a huge capacitor This capacitor is connected in parallel over about 45% (R2) of the total resistance It is this capacitor in parallel over a part of the resistive path that constitutes the difference between AC and DC measurements! R1~55% C R2~45% This is the generally accepted simplified circuit equivalent circuit of a lead acid cell

AC Measurements The Ohmic value of a capacitor depends on the size of the capacitor and the frequency of the test current. XC=1/2fc The higher the test current frequency and The bigger the capacitor, the smaller the ohmic value of the capacitor. The typical capacitance value of a Cell is 1.5 Farads per 100 A-hr of capacity.   The typical test frequencies used by the present test equipment manufacturers ranges from 20 Hz to 1000 Hz.

AC Measurements When applying an AC test current over this circuit, the capacitor becomes a partial short circuit around part of the resistance path. The applied AC test current signal will follow the path of least resistance and therefore mask rising problems in the important R2 path R1~55% R2~45% C As the ohmic value of the capacitor approaches values that are close to 60% of the baseline value of R2 it becomes practically impossible to detect any increases in the R2 part of the path

Typical Cell problem, detected by increased resistance. Paste breaking away from grid causing increased resistance primarily in the R2 path. This picture shows a positive lead calcium plate that is approximately 15 years old. Note that the grid has grown due to corrosion, which takes place gradually as the battery ages. The paste pellets are breaking away from the grid structure, causing a high resistance connection between the grid and the paste. New Grid Aged Grid

Circuit Analysis R1 = 110μΩ C = 15F R2 = 90μΩ We will use the below formula to calculate the Impedance if the resistance values in R1 respectively R2 changes R1 = 110μΩ C = 15F R2 = 90μΩ   So now let us analyze the difference between AC measurements and the true Internal Resistance. Guideline for Capacitance is 1.5F / 100Ah. Example 2000Ah = 30F This example shows typical values of Baseline Resistance and Capacitance for a 1000 A-hr Cell. Using this model, we will examine what the impedance of the circuit is for the values shown and then we will calculate what the impedance readings are for different failure modes. We will assume that R2 increases enough to cause the total resistance to increase by first 25% and then 50%. We will then see what happens as R1 fails Typical 1000 Ah cell

% Change Rtot from baseline % Change Ztot from baseline R2 vs. Rtotal and Ztotal Test Freq Cell failure Rtot R1+R2 % Change Rtot from baseline Ztot % Change Ztot from baseline 60 None 200  185  R2 > 140 250  25 208  12 R2 > 190 300  50 220  19 200 139  135  -2.2 133  -4.0 The analysis comes up with some interesting answers; Note that the Impedance of a known good cell always reads less than the Resistance. Also note that if the total Resistance increases by 25% as a result of R2 increasing, the corresponding increase in Impedance at 60 Hz is only 12%. The change in Impedance at 200 Hz is actually a decrease.   This shows that Test Equipment or Battery Monitors using these frequencies will actually declare a bad cell good. The problem with using Impedance as a State of Health Indicator gets even worse if the 50% above baseline criteria is used.

% Change Rtot from baseline % Change Ztot from baseline R1 vs. Rtotal and Ztotal Test Freq Cell failure Rtot R1+R2 % Change Rtot from baseline Ztot % Change Ztot from baseline 60 None 200  185  R1 > 160 250  25 234  26.5 R1 > 210 300  50 284  53.5 200 139  187  35 236  70   This table shows what happens if the total resistance increases as a result of R1 increasing. At 60 Hz there is only a small error introduced, but at 200 Hz the Impedance reading goes way high and exaggerates the problem. This means that using a 200 Hz device will cause good cells to be declared bad well before they reach their Resistance failure level.

Noise Typical AC Ripple riding on a 2 volt UPS Cell Ripple voltage is ~40mV or 40,000µV. Most AC instruments inject a 1 A AC test signal which will generate a 300µV signal through a 300µΩ cell This means that the test instrument will have to accurately resolve a 300µV signal in 40,000µV ripple noise! 1A * 0.0003Ω=0.0003V This is an oscilloscope presentation of the ripple voltage present on a 2 volt UPS battery with a baseline internal resistance of 300 micro-ohms. Note that the ripple voltage is around 40 millivolts peak to peak that is 40,000 micro-volts. A 1 amp square wave AC instrument would only generate a peak to peak voltage of 300 micro-volts. That gives you an idea of the signal to noise ratio.

Resolution/Test current Example: The Resistance of a 1000 Ah cell is 200 . 25% increase of resistance designates a bad cell which means that the test instrument has to detect an increase of 50 to 250. In order to accurately measure this increase, the instrument needs a resolution of 10% of the measured value. This means that a 1 amp test instrument will have to resolve 5 V to detect a cell turning bad! That’s a very expensive voltmeter… Low current instrument may yield reasonable readings on very small high resistance batteries such as 12 volt modules, but cannot realistically be used on large 2 volt cells. 1000Ah = 200microohm + 25% 250microohm , 50mocroohm over 1 A equals 50microvolts. In order to measure 50microvolts accurately a voltmeter will need the capacity to resolve 5microvolts. Standard Fluke voltmeters can typically only measure down to 0.1mV.

How are Resistance Measurements made? The instantaneous voltage drop at time zero and when the load is removed shows the voltage drop across the internal resistor Resistance = V/I So if we measure the test current flowing and the corresponding change in voltage, then we can simply solve ohms law to calculate the internal resistance.

Ohmic measurements Alber’s Internal Resistance measurement method is superior because of the following reasons. Eliminates the “capacitor effect” Not affected by ripple or noise superior resolution and repeatability These are the three different terms used in connection with Ohmic measurements. Please note that AC Impedance should be simply called Impedance. The definition of Impedance is opposition to AC current flow. Therefore AC Impedance is redundant AC Conductance is not correct; the definition of Conductance is the inverse of Resistance. The commercially available AC Conductance meter actually measures the same thing as Impedance except it inverts the answer. In this presentation it will be treated as an Impedance meter.

Case Study How does a failing cell look like and how can we detect and prevent a catastrophic failure? The following case study Illustrates how the Alber monitor detects a failing cell and how a discharge test confirm the internal resistance reading.

Playback The following slides show an internal resistance test and then an actual 20 min battery discharge with the playback accelerated It is a battery string consisting of 60 - 8 volt jars. One of the cells in jar 41 fails. The real time display ability during the discharge confirms the resistance test results.

This is an indication of a bad cell within this jar. The 941 micro-ohm value of Jar 41 is 22% higher than the other jars in the string and has triggered an alarm turning the bar red. This is the results from the automatic resistance test. It’s a string consisting of 60 8-volt jars. This is an indication of a bad cell within this jar.

Now we will step though the discharge.

At 13 min jar 41 begins to fall off.

At 17 minutes jar 41 is 4.99 volts and the test is aborted.

Note that jar 41 is now floating with rest of the jars. By 18 minutes, the charger has walked back in and there is 90 amps of charge current. Note that jar 41 is now floating with rest of the jars. Without the internal resistance alarm, real time capture and storage, this failing cell would have been overlooked and left in service only to fail during an outage!

How? Confirming the data We verified the data from the resistance test in a discharge test. Knowing how each individual cell performs during a discharge can be crucial. High scan rate is important

Auto Discharge Capture How? Most power outages that occur are less than 30 seconds. Most monitors cannot record cell voltages in these short events because their scan rate is too slow. This prevent them from capturing discharge events which provides important information.

Remember: A battery is only as good as its weakest cell. Conclusion Make sure you select a monitor that will provide you with correct information Correct and accurate parameters as described in this presentation Auto capture of discharge events Real time display with high scan rate (<10 sec) Make sure your maintenance crew attend a Basic Battery Training Remember: A battery is only as good as its weakest cell.

Battery Monitors BDS-256 Battery Diagnostic System UL Certified CE Approved Monitor any battery system up to 600 volts DC UPS systems Generating stations Industrial

Parameters Monitored Overall voltage Discharge current Charger float current (optional) 2v cells, NiCad cells, 4v, 6v, 8v and 12v modules Temperature Resistance of all cell/jars, intercells, and intertiers

System Level I/O System inputs System outputs (form C contacts) Remote alarm reset 16 digital inputs System outputs (form C contacts) Maintenance alarm Critical alarm 8 programmable control outputs(BDS-256)

Communication Two RS-232 for local computer RJ-11 for telco dial up RJ-45 for Ethernet connection Standard Modbus protocol

BDS Controller The “Brain” that control the system. Collects and stores data from the DCMs Microprocessor driven Stand alone – No on site PC required. 8 strings of 256 cells per Controller

Data Collection Module (DCM) The Data Collection Module acquires all readings from the battery 48 cells or modules 2 temperatures Discharge current Charger float current

External Load Module Supports proactive DC internal resistance testing One ELM for each string Tests battery in 10% increments Test current approximately 30 amps

Configuration examples 3 strings, 40 jars, 12V per jar 1 DCM and 1 ELM per string 1 CT and at least 2 temperature probes per string 1 controller String 1 String 2 String 3

Configuration examples 2 strings, 240 jars, 2V per jar 5 DCM’s and 1 ELM per string 1 CT and at least 2 temperature probes per string 1 controller String 1 Fiber optics for data transfer and power supply between DCM and Controller String 2

Ease of Installation Modular design allows the DCMs to be located near the battery, reducing wire lengths One 120 vac power connection required for up to 8 strings of 256 cells.

BDS-256 Installation Mounted on top of MGE 6000 Cabinet

MPM-100 Designed for applications below 130VDC UL and CE approved More than 100 battery configurations available Monitors 1 string of 120VDC or up to 4 strings of 12, 24 or 48VDC Modbus protocol Powered from DC bus or 115VAC Network, Serial, Modem connection options

Alber the total solution

MPM installation

Thank you Press ESC to exit

Pipe analogy A healthy battery will produce power and allow easy flow of DC current. The more power that is required, the more power will be produced When the fuel is gone, it has to be charged

Aging As the battery ages, the capacity diminishes. It is as if the pipe has clogged up. It cannot produce the desired capacity

AC based testing A battery has a huge built-in capacitor (Parallel plates) Capacitor will allow AC current to flow but will block DC current When testing with AC the battery may look healthy as the AC test current will pass through capacitor

DC based Testing Alber’s DC test does not look at the AC path It measure the battery’s resistance under normal working condition This makes it possible to assess the health and detect early signs of degradation

Test current Ohm law (V=I*R) dictates that low current over a low resistance will result in low voltage. Low test current will not allow for accurate measurements. Resolution above +/-100microOhms will not detect failing cells with more than 500Ah capacity

Test current The Alber monitor use 30A-70A of current to allow for an accurate reading. The Alber test method is repeatable and accurate. Alber internal resistance test equipment can test any battery design in increments of 1µΩ

Noise The UPS battery is a noisy environment Using low test current is like a conversation with someone with low voice in a noisy bar Would you choose that environment for a discussion where attention to details are vital?

Return