Solutions
   
Threat Detection & Chemical Monitoring
 

Vertical Axis, Dual Blade Mag Lev Wind Turbine (con't)

Power Distribution: The power will be distributed to the BCM, SCM and SIOC units via a power distribution network which will run alongside the data communications cable. For the final, field deployed system, an integrated cable combining power wires and the data network wires will be used.

Power Consumption: The decision to use a single power source for entire system was based on ease of maintenance. One chargeable battery should be capable of powering a system for over 5 years without maintenance. Power consumption by the SIOC, SCM and BCM units is a function of the frequency of usage (traffic on the bridge). The system will be designed to go into sleep mode when there are no tasks to be performed. Whenever traffic is on the bridge, the system will wake up to start monitoring and sampling the sensor data. Power consumption will reach peak only when the system is in use. At this stage, estimations are made on assumptions of usage by the SCM and BCM. In Phase II, a re-chargeable 12 volt battery will be used. When the system is tested on a bridge in the CFC-WVU lab, the constants will be determined for the power consumption model. The following formula will be used to calculate the power consumption (Pt):

where Pt = Pi * Zi + ( 1 – Pi)*Za
PT = total power consumed
Pi = power consumption in idle state
Zi = probability of node in sleep mode.
Za = Probability of node active

For Phase II, a 6-12TLA stationary modular battery (Fig 5.x) from AGM Technology will be used and integrated with a 12V, 700mA, portable briefcase solar generator To trickle charge the battery(Fig 5.x). The battery will be installed in the BCM’s enclosure. Along with the battery, the enclosure will include electronic interface between the solar panels and the battery, to ensure the battery is not damaged by over-charging.

Designing a power supply system for a bridge monitoring application requires knowing the power consumption, current draw rate, and the anticipated maintenance cycle duration and deployment environment (such as the safety of solar panel against theft), and the anticipated battery life (such as 5 to 10 years without maintenance).

Each hardware component will consume 2 to 5 watts of power and will require 2 amps (peak) of current. The goal will be to provide a minimum of 5 years of uninterrupted, regenerative power to the system. This will be achieved using two options (both options would provide the same amount of power):

  • Solar panels: In safe and stable areas, solar power panels will be implemented. These panels are inexpensive; however, they can be easily stolen or targeted.
  • Fuel cell batteries: In where environments where tampering and targeting are issues, low cost fuel cell batteries could be deployed. Their advantage is that they can be hidden in the bridge structure itself. They are already in limited use by the DOD. The disadvantage is that they are more expensive than solar panels.

A wide variety of options for solar and fuel cells will be reviewed in Phase II: from simple sealed lead acid battery packs to complex Lithium-ion packs that include electronic safety, monitoring, and charge-control circuitry. Critical issues, such as cost, packaging, eco-environmental factors, and regulations affecting battery-pack design and construction will be considered, as well as bridge parameters, like bridge length, number of bridge sections, number of sensors, and anticipated traffic volume.

Solar Power: Solar panels are one option for recharging the batteries. The solar charging system will be rated for 12 volts and a minimum of 20 watts of power, which is sufficient for recharging any battery. In Phase I detailed analysis of the availability of solar chargers was conducted. The conclusion reached was that solar is the most economical and practical option for this application; however, its disadvantage is that solar panels can be easily stolen or used as a target, which would render the system useless without power. The solar configurations would include a solar panel installed on a pole, which could also serve as an antenna. (Figure 1) is a diagram of solar backup system includes a solar panel that charges a battery. The solar panel is connected to a regulator to protect against overcharging and damaging the battery.

Battery Chemistry: Today, the choice is typically one of three rechargeable cell types: Nickel-metal hydride (NiMH), Lithium-ion (Li ion), and Lithium-polymer (Li-polymer). Li-polymer’s advantages are higher energy density by weight than Li ion and higher volumetric energy density in thin formats using less than 5mm cell thickness. Unlike other chemistries that are typically available in limited standard sizes, Li-polymer is available in any footprint, which will provide greater flexibility during design. Li-polymer’s stability in over-voltage and high temperature conditions provides a wider margin of safety than with Li-ion. Li-They are also weatherproof.

Fuel Cell Battery: An electrochemical cell in which the energy of a reaction between a fuel, such as liquid hydrogen, and an oxidant, such as liquid oxygen, is converted directly and continuously into electrical energy. In addition to this pure hydrogen type, there are hydrocarbon fuels for fuel cells, including diesel, methanol and chemical hydrides. The waste products with these types of fuel are carbon dioxide and water. The materials used in fuel cells differ by type. In a typical membrane electrode assembly (MEA), the electrode–bipolar plates are made of metal, nickel or carbon nanotubes, and are coated with a catalyst (like platinum, nano iron powders or palladium) for higher efficiency. Carbon paper separates them from the electrolyte, which could be ceramic or a membrane. UltraCell Corporation of Livermore, CA offers a 12 volt fuel cell (Figure 2) that provides 12 volts power with total energy capacity (per charge) of 180W.

The sensor networks will be designed for deployment in adverse and remote/non-accessible areas that may not have fixed infrastructures. Power harvesting technology extracts energy from relatively inexhaustible ambient sources, such as wind, temperature, vibration and other natural phenomena, to provide power to the sensor nodes. Additionally, energy usage optimization will be employed, based on previous work developed by Erallo for a MEMS-based sensor application. A number of power harnessing technologies will be reviewed and considered for this application; including solar and vibration energy. Harnessing an ambient energy sources like vibration may even eliminate the need for batteries in many monitoring applications. This would extend the useful life of the sensor system and significantly reduce the lifetime cost of the sensors.

Vibration Energy Harvesters (VEH): The two most widely studied classes of vibration energy harvesters (VEH) are inductive and piezoelectric. Inductive VEHs work on the principle of the magnetic generator of Faraday’s law, with the external input energy coming from the ambient vibrations. The vibrations generate a displacement between a permanent magnet and a pickup coil, generating voltage in the coil. The voltage is rectified and delivered either to a storage device or directly to a load.

Inductive VEHs are well suited for use with lower frequency vibrations (below a few hundred Hz). Inductive VEHs are robust and can survive significant temperature and shock extremes and are ideal for battlefield environment. Inductive devices can last for decades with little or no degradation in performance.

Vibration based power harnessing: Our adviser for this technology will be Dr. Robert O'Handley, Chief Scientist and Co-Founder of Ferro solutions. Ferro has developed a suitable vibration generator (Figure 3), called VEH-3. These units will be used in experiments to analyze their suitability for BMS applications. These units are currently being used by the Navy to power sensor networks in ships and submarines. Ferro Solutions has a plans to modify and fine tune the generator in Phase II, if required.

Solar power: Self-contained, portable solar power generation systems will be considered as an alternative to vibration power generation. These devices are built like a generator, with everything except the 50-watt solar panel contained in one attractive enclosure. These solar devices can power a 12 volt battery charger and can provide 50 Watts of power. This would be sufficient for powering sensors as well as the BCM. A detailed analysis of the daily power draw from the sensor module and BCM module will be conducted so that an appropriate solar power unit can be identified and analyzed.

Power storage: Energy storage is comprised of a group of elements used to buffer the energy coming from the power generator (Figure 4) and deliver them to the mote in a predictable fashion. Designing the energy storage involves choosing the storage elements and charging mechanism for correct operation and efficient energy transfer while satisfying a set of system requirements such as lifetime, capacity, current draw, size and weight. For the energy storage element, NiMH (Nickel Metal Hydride) or Li+/Li-polymer (Lithium-ion / Lithium polymer) batteries are desirable due to their high energy density while super capacitors are desirable for their high charge cycles.

 


 


 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1 Photo of a solar panel recharger and diagram illustrating a solar backup battery system

 

 

 

Figure 2 UltraCell Fuel Cell Cartridge

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3 Ferro Solutions’ Vibration Power Generator

 

 

 

Figure 4 Solar Power Generator

Erallo Technologies is an 8(a) SBA-certified company. © 2012 Erallo Technologies Inc.. All Rights Reserved.