Published in engi.philica.com
In theory, with the grid connected and when an induction machine’s speed increases beyond its synchronous RPM (Revolution Per Minute), its function changes from a motor to a generator. Slow devices such as wind turbines or waterwheels for harnessing renewable energy employ speed-increasing gearboxes. This paper presents several electronic approaches to operate those machines in generator mode, eliminating the need for gearboxes.
Since the invention of the induction motor in the late 19th century, it has been evident that when a motor is driven 7% to10% faster than its synchronous RPM it becomes a grid-tie induction generator contributing power back to the grid. Also, because over the latter half of the last century oil has become a relatively rare and expensive commodity, we are turning to renewable energy for an inexhaustible, abundant and widely distributed resource. Among these renewable sources, kinetic power from water and wind current may be the most popular. Because these resources are widespread geographically, the potential is abundant; however, the devices that harness them are slow and require gearbox, the whole system is not only a high maintenance component that is prone to failure, but is also expensive, constituting as much as 50% of the total cost of the system. Thus, there is great need for robust, simple and inexpensive replacement solutions.
This paper presents several methods of operating an induction machine in generator mode that avoid a mechanical approach, which would require driving the machine above its synchronous RPM speed. The methods enable an induction machine to function as a grid-tie induction generator at very low speed, from a few RPM to 30% of its synchronous RPM.
Over the course of decade-long endeavor, each version of our design became more streamlined, simpler and "meaner" than the previous one. In retrospect, one may find that the work looks too simple and wonder why it was not conceived before. Although more research may be required to fully explain the phenomena described here, it is not the purpose of this paper to provide, elaborate theoretical discussion. Here, we present reproducible facts: the designs and their results.
This paper calls upon multiple areas of discipline including electronic and electrical engineering. To facilitate the discussion, it is worthwhile to introduce the components used in the designs and their relevant characteristics.
Figure 1 - Diode
Diode (Figure 1): This is a two-terminal device with an anode and a cathode. It only allows current to flow when the voltage of the anode relative to the cathode is positive. One may use a diode to access either the positive half or the negative half of an alternating current (AC), but not both.
Figure 2 - SCR
SCR (Figure 2): This is a three-terminal DC device with an anode, a cathode and a gate. Although it may be used for other purposes, it has a switching mechanism that allows a DC anode-to-cathode flow when the gate-to-cathode current becomes positive.
Figure 3 Triac
TRIAC (Figure 3): This device has three terminals: MT1, MT2 and gate. It functions as a AC electronic switch or relay allowing the current to conduct when a small current is applied to its gate. It should be noted that this device is designed for AC, and therefore the polarity of its currents are inconsequential. Its main terminals, MT1 and MT2, conduct in either direction when a current is applied to its gate regardless of whether the current is positive or negative.
Dv/dt: Gate-controlled electronic devices such as the TRIAC or SRC may inappropriately become conductive when a sudden change or surge occurs in voltage difference between its main terminals. The term used to describe this phenomenon is dv/dt. In practice, one may use snubbering circuit to prevent these unexpected faults to happen or, in our case, on place of each TRIAC or SRC, we use a common, joined gate, serially connected pair.
Figure 4 - AC Grid
The AC power grid (Figure 4): Alternating current is the form in which electric power is delivered to businesses and residences. Its usual waveform is a sine wave. For the purpose of our discussion, we assume it is single-phase with one "hot" wire carrying a sinusoidal alternating current and a ground wire. A 3-phase system may be considered to contain three separate single phases, all sharing the same ground fault wire.
Figure 5 - Motor/Generator
The Motor/Generator (Figure 5): For the purpose of our discussion, we assume the motor is a multiple pole, single phase, squirrel-cage induction motor without any starting mechanism; if there is one, it should be disabled or removed. The industry has several standards for motor specifications in different regions of the world; here, the NEMA standard, which is widely used in North America, is assumed. Each motor/generator has a manufacturer-specified synchronous RPM with the formula:
n = 120*f / p
Where n is the synchronous speed of the AC motor in RPM, f is the power supply frequency, and p is the number of poles.
The systems provide the induction machine with rectified DC, which is the result when AC is converted to DC. Upon receiving the DC, the armatures become electrical magnets. When the rotor turns, the aluminum bars inside the rotor act as coil cutting through the magnetic field. The device now operates in generator mode; however, the AC to DC conversion has to provide a means of reverse bias, inverting the now-amplified DC back to AC to complete its grid tie functionality.
The machine converts mechanical power to electricity only when it is driven by an external prime mover torque, without which the system is only a short circuit generating heat according to Ohm's law. Therefore, it is imperative to take measures to prevent the generator from becoming engaged with, or disengaged from, the grid when the generator is stationary. The activation or deactivation of power from the generator to the power grid AC should only be performed when the system is running; a recommended engaging speed is between 1%-10% of the synchronous RPM.
Our development began with relatively complex models, but as each generation passed, the designs became simpler; however, here we present the models in the reverse order. This may be why the simple solutions have not been discovered and implemented until now.
Half-wave diode-base module
Figure 6, Figure 7 - Haft Wave Modules
This is the simplest level of implementation, in which the motor/generator and a single diode are serially connected to the grid (Figures 6 and 7). Please note that the diodes are unconventionally placed between the motor/generator and the ground terminal of the AC grid.
How the system works
- For this discussion we examine the circuit in Figure 7, in which the grid supplies positive current to the motor/generator. This is not a constant current but a variable DC.
- The variation of this current energizes the armature with a variable flux magnetic field, and if the rotor is in motion, then the aluminum bars inside the rotor form closed coil cutting through the magnetic field to induce a current.
- Because the supplied flux is variable, this newly formed current is also variable. From the construction of the induction motor, this current adds onto the supply current; therefore, the generator now acts as a current amplifier, creating a voltage elevation.
- When the "hot" leg of the grid is connected directly to the motor, any added current is allowed to flow back to the AC power line, thus creating a grid tie operation.
Figure 8, Figure 9 - Oscilloscope Images.
One may wonder if there is any change to the AC power curve before and while the generator is in operation. Figures 8 and 9 are redrawn from captured images from an oscilloscope. Indeed, during the generator mode, the voltage between its two input terminals increases by as much as 50%.
Figure 10 - Output Curve.
*XX= 8.5 the rate output of the motor.
Please see Appendix B for the setup and data table.
Our equipments are only capable to drive the generator at 1% its synchronous RPM; (~18 RPM) and above.
The power curve appears as shown in Figure 10 with +/- 5% accuracy at 119v AC input. Please see the below note 4.
The manufacturer attaches a nameplate to the motor showing its specifications, including power (usually in horsepower), RPM, and number of poles. It is assumed that when one operates the motor in motor mode, its capacity is as specified. One would also expect that under normal conditions the output when the motor is in generator mode should be approximately 100% of the specification. Yet, Figure 10 shows an 8.5-fold increase from the rated power. The reason for this phenomenon is unknown, though there are several known factors that affect a motor’s operation.
In normal operation:
a- Every time the AC changes its polarity, power is spent to re-energize every pole of the armature to the opposite polarity.
b- Only during this brief period of the AC curve does the motor actually produce torque.
c- Although the rotor is in motion, when the armature changes its polarity, this also causes the rotor to change its polarity.
d- In a 3-phase system, each phase creates its own magnetic field at the armature. The fields interact with each other to pull and push, and the difference in intensity causes the torque.
In our case:
a- Because we power the machine with DC, the armature does not change its polarity.
b- The motion-to-electricity conversion happens throughout the whole course of the active curve of the AC. In this specific model, the conversion occurs 50% of the time when the AC is either positive or negative.
c- Later in this paper, full wave models will produce even greater results.
For a long time it also puzzled us that because in this model we only set the AC to be active during half of each cycle, the oscilloscope data should show a change only during these half-cycles. After extensive observation, we realized that the minor changes in these curves, they depend on the motor's distance from the main breaker or main transformer at the utility post. We then realized that the transformer indeed resemblances the curves to a more balanced fashion.
At normal operation, a generator produces its full output when it reaches at around 107% its synchronous RPM, in our settings, the machine produces its peak, full output at around 1% its synchronous RPM. And its rate slowly but steadily reduces when it's driven faster; our measurements show that its output decreases about 9% when its speed reaches 27% the synchronous RPM.
We are also able to observe that the output is somehow proportional to the square of the supplied voltage, therefore slight variations on the voltage result much larger output swings. Please see appendix C.
2- Double half-wave diode base module
Figure 11 Double Haft Wave Model
The above scheme works; however, in our opinion it will achieve a balance if it is implemented in two separate, back-to-back diode circuits to drive two identical motors or, in our case, two circuits in the same housing (Figure 11). Additionally, the figures in Appendix A show how a 3-phase generator was converted into a single housing with two separate "generators."
3- Full-wave TRIAC or SCR modules
Figure 12 Triac Module
Figure 13 - SCR Module
In our earliest development work, we started with full wave modules. Although these modules may seem out of date or irrelevant, they are still in our arsenal as a part of a variable-speed automated system allowing full control of its speed and the amount of electricity it harnesses. There are two modules: a simpler one utilizing TRIACs (Figure 12) and a second one using SCRs and diodes; these are intended to operate ½ HP motors for the US 110v system or the EU 220v system, respectively. Each has its own advantages and disadvantages. The purpose of this section is to present alternate solutions to operate an induction generator in its full wave configuration.
Please note that if one chooses to implement this scheme, professional help may be required to make sure that the system works well with a specific generator and does not suffer from false triggering due to the characteristic dv/dt of the selected TRIAC and SCR electronic components.
Figure 14 - Output Curve
*XX= 24.5 the rate output of the4 motor.
Our equipments are only capable to drive the generator at 1% its synchronous RPM; (~18 RPM) and above.
The power curve appears as shown in Figure 14 with +/- 5% accuracy, at 119v AC input.. The graph is plotted from the operation of an SCR module. Please see appendix D.
White paper topics
We devote this section to the discussion of a variety of issues related to the application of this new type of device.
The data show that the generators operate at ten times the power specified on the motor's nameplate. At low motor speed, with such high power, the built-in fan becomes useless. Another measure, such as an extra fan, must be in place to remove heat from the generator. Moreover, higher outputs imply higher currents, resulting in much more heat than normal. One should consider either adding extra radiating surface area or using a liquid cooling system.
Three-phase systems are designed to allow the magnetic fields of each phase to interact in order to provide a smoother torque when operating as a motor. When operating as a low speed generator, however, such interaction is counterproductive. Although it is possible to operate a 3 phase generator; in that case, a rearrangement of the position of the poles is needed to counter those unwanted reciprocal torques. An alternate method could be to use 3 separate generators (possibly in the same housing); or in the case of multiple installations, such as in a wind farm, using each installation to generate a phase may be an even better, simpler approach.
As with any other inductive device, such as a motor, transformer or generator, especially in a high capacity installation, inrush current is an issue that must be addressed.
The motors / generator.
When motor leave the factory, the capacity of their wiring is rated for the stated input or output as near the maximum tolerance. When one decides to operate a certain motor as generator; he/she should ensure that they are working within the tolerance of the hardware and for the prolong use.
Switches / Relays.
It's very common in automate systems to use gate driven, solid-state switches or relays to engage or disengage to the grid; they may or may not work as such. Normally a electrical system has two sides the source to provide the energy and the load to consume it. With our systems, the roles are interchanging thus may cause problem to such gate driven components.
Although these are breakthrough findings, they do not require a high level of expertise to implement. All the necessary components are widely available as off-the-shelf items. The findings are not intended to harness energy but to enable energy-harvesting devices to produce the kind of electricity that is widely used, i.e., to perform a grid tie. The approach is simple, robust and inexpensive, and it opens a door to areas that were previously inaccessible. History has proven that with human ingenuity, once a door is opened the horizon is the limit.
According to NEMA standards, a 12-lead, 3-phase, ~1800 RPM induction motor has its poles and associated wiring distributed as shown in Figure 15. These instructions are intended for a knowledgeable audience with sufficient expertise to understand and use this as an illustration of how to set up a motor of a specific size and output. As an example of the type of knowledge needed, our instructions assume that all of the poles are to be serially connected. This fact by itself may reduce the power of the motor compared to the amount appearing on its nameplate.
Figure 15: Lead distribution.
The circuit in Figure 11 has two motors (labeled "M"). We will assign one machine to drive the first motor's poles having the following leads: (1, 4), (2, 5) and (3, 6).
Our intention is for the first motor's (1, 4) pole to have positive polarity and the other two poles (3,6) and 2,5) to have negative polarity.
Similarly, the other motor's (8, 11) and (9, 12) poles will have positive polarity and the (7, 10) pole will have negative polarity.
One may use low-voltage rectified DC to test each pole with the help of a compass.
One might also use the following instructions, though they require opening up the motor for verification with a compass.
Step 1: Connect the 4 following pairs: leads 4 + 5, leads 2 + 6, leads 9 + 10 and leads 7 + 8.
Step 2: The other 4 leads; 1, 3, 11 and 12; are connected to the system as illustrated in Figure 16.
Step 3: Open the motor, remove the rotor, and place it on its side as if the rotor is standing upside down.
Step 4: Apply a low-voltage AC to verify its configuration (e.g., with a compass), as shown in Figure 17.
Figure 17 - Polarity distribution.
Step 5: Put all the parts back together and the system is ready to operate.
The following table is captured with a 4 pole 240V ½ HP, 119v input.
OUTPUT in watt
A 4 pole 240V motor produces:
141 Watts at 55V input.
591 Watt at 110V input.
With a 4pole 240v ½ HP motor, at 119v input, at 18 RPM outputs 2351 W.
Information about this Article
This Article has not yet been peer-reviewed
This Article was published on 13th February, 2012 at 23:41:31 and has been viewed 14000 times.
This work is licensed under a Creative Commons Attribution 2.5 License.
The full citation for this Article is:|
Tran, P. (2012). Electronic approaches to direct drive an induction generator; without mechanical gearbox.. PHILICA.COM Article number 316.
1 added 10th July, 2015 at 19:17:53
The article discuss the latest findings. However there is other way to do.
- The easiest mean is to operate the motor at 1/5 its intended voltage.
- The second method is to employ a variable transformer or a step down (1->1/5) transformer.
- The method that use variable transformer allows it to be adjustable.
And finally the method to employ capacitor as current limiting device.