Flirting with disaster

Network expansion keeps altering power needs, creating distributed networks that require back-up batteries in harsh environments. Under extreme conditions, the batteries' thermal properties may be insufficient maintain stability. This can cause thermal runaway, a disastrous condition that can result in hydrogen gas explosions and meltdowns of battery cases-obvious threats to the continuity of telecommunications service.

However, aggressive thermal management may be enough to prevent a thermal runaway event.

The intrinsic problem Batteries used as back-up network power sources are float-charged, with a constant voltage continuously applied to the battery string. The float voltage is always high enough to cause a small amount of current to flow through the battery. Depending on battery chemistry and design, the flow current consists of different components, such as the electrolyte electrolysis or internal gas recombination.

The float current has two important properties: It heats the battery internally, and it is sensitive to the ambient conditions at a deployment site.

Generally, the float current increases with rising float voltage and ambient temperature. This means that the amount of heat generated inside the battery can vary with time and can be affected by external factors. Under normal conditions, the amount of heat deposited inside a floating battery is small. It results in almost undetectable increase of the battery temperature.

However, at higher temperatures or float voltages, the amount of deposited heat increases. This by itself does not mean that that the battery is in trouble. If the battery can dissipate the internally generated heat, the system is in thermal equilibrium with the environment.

Unfortunately, it is possible to reach conditions at which the battery can no longer dissipate all the heat. When this occurs, the internal battery temperature will rise, which will result in higher float current. This will continue until the rectifiers are unable to supply the current demanded by the battery or until the battery fails (meltdowns of battery cases have been observed).

Called thermal runaway, this phenomenon occurs because the battery can no longer dissipate the internally generated heat, thus increasing internal temperature.

The telecom power community is most familiar with thermal runaway events observed in valve-regulated lead acid (VRLA) batteries, but thermal runaway is not limited to VRLA batteries. It can also occur in other battery systems. The recombination feature of a battery, such as the internal oxygen cycle in VRLAs, contributes to its susceptibility to thermal runaway.

A common misconception about thermal runaway is that it can occur only at high ambient temperatures. However, it also can be triggered at low temperatures if the battery float voltage is high enough.

The thermal runaway events observed in aqueous batteries normally are accompanied by evolution of hydrogen. The rate of this evolution process increases as the thermal runaway progresses and may result in the hydrogen/oxygen mixture exploding inside the power node, destroying the enclosure.

The high temperatures inside the battery cells during the thermal runaway events also may generate poisonous and corrosive gases (such as hydrogen sulfide and sulfur dioxide). These gases are released to the enclosure environment, and not only destroy equipment but also create a health hazard to carrier personnel.

Susceptibility of a battery system to thermal runaway is usually a function of two parameters: effective float voltage per cell and battery temperature. The higher these parameters, the more likely the battery will undergo thermal runaway. Increased float voltage per cell can result from faulty or incorrectly set rectifiers or from the shorted cells in the battery string.

High battery temperatures usually are caused by the environment and consist of several elements, such as the outside temperature, solar loading, internal heat sources such as rectifiers, and thermal management inside the power node enclosure. A failure in the thermal management system may result in significant temperature increase inside the enclosure.

A progressive disorder The most obvious hallmarks of thermal runaway are its common disastrous after-effects. However, runaway also can be detected well before catastrophes occur.

The two most obvious indicators of thermal runaway in progress are the battery temperature and the float current. Both are direct results of the system's inability to dissipate the generated heat. The difference between the battery temperature and the ambient temperature, is sufficient to detect the thermal runway condition.

A simple way of using the temperature difference is to set a "thermal runaway threshold." When the temperature difference crosses the preset temperature (for example, 15 C), the alarm is triggered.

More sophisticated methods of tracking a battery's thermal state rely on continuously following and analyzing temperature changes in the ambient environment and the battery, and the difference between these two parameters. They can also take advantage of monitoring the changes in float current.

The methods of detecting thermal runaway conditions are the first steps in preventing it from occurring. The sooner the thermal runaway event is detected, the better the chances are of controlling it without the loss of service. Because thermal runaways can have many different causes, responses will vary based on cause, especially when the process is detected early.

However, quite often, the only safe way to handle thermal runaway in progress is to disconnect the batteries from all sources of electrical power: the rectifiers and other battery strings. The nature of thermal runaway in today's telecom batteries requires a source of external power for the phenomenon to proceed. By cutting off the power, the electrochemical reactions responsible for the current flowing through the battery-and the generated heat-are stopped.

The occurrence of thermal runaway in a power node is serious, and stopping it does not mean that the problem is solved. The actual causes of each thermal runaway should be thoroughly investigated, and the lessons learned from those investigations should be implemented in the field to prevent future events.

Preventative planning Thermal runaway occurs because the batteries become overwhelmed by radical changes in their overall environment. Discovering such conditions before field deployment can help prevent thermal runaway.

The system modeling techniques can help predict possible conditions in the network's power nodes. These conditions should be compared with the thermal safety envelope of the candidate battery. This envelope is a set of conditions (pairs of ambient temperature and float voltage) that do not result in thermal runaway.

In the past, the method used to determine the possibility of thermal runaway in a battery was to perform single temperature/voltage combination tests. The float voltage in those tests corresponded to a fixed number of shorted cells in a telecom battery string. The test was performed at an elevated temperature.

This pass/fail test was based on whether or not thermal runaway was observed. Such tests yielded limited information. If thermal runaway was not observed, the system was thermally stable at the particular combination of float voltage and temperature.

However, the test did not tell how much heat and voltage the tested battery could accept before going into thermal runaway. Similarly, if thermal runaway was observed, the only information available was that the tested battery was not thermally safe under the test conditions. This was a trial-and-error approach. It was time consuming, and it would require many battery units for determining the safety envelope.

The new method for testing batteries' thermal properties simplifies the process by very slowly changing one of the test parameters until thermal runaway is observed. Tests are performed at constant ambient temperature, and battery voltage is varied. The test starts at a normal battery float voltage, and after the battery has reached a constant temperature, the applied voltage is slowly increased. The rate of voltage increase must be low enough to allow the battery to reach thermal equilibrium with the environment. Under these conditions, thermal runaway can be detected by the relationship of the ambient temperature to the applied voltage.

The test is no longer pass or fail. It provides the information about the conditions present when the thermal runaway phenomenon was actually observed. To determine the thermal safety envelope of a battery, the test should be repeated at different temperatures. The temperature range should be between the lowest and the highest temperatures expected to be experienced by the batteries. The results of the tests can be conveniently summarized in a plot of voltage and temperature values that result in thermal runaway (Figure 1).

The new testing method greatly reduces the number of thermal runaway tests necessary to determine the safety envelope of a particular battery.

The potential for thermal runaway may be an inherent risk of network deployment, but if conditions are monitored and batteries proactively tested, it can remain a risk never fully realized.