2.1_ Description of the proposed system
Here is an exploded diagram showing the core of the heating system, constituted by the windmill and the water-stirrer:
The vertical shaft actuates the water-stirrer down-below. This stirrer is housed within the column constituting both the main water-tank and the backbone of the building.
A variable quantity of water is pumped into the stirrer from the water-tank down-below. The resistive torque is thus controlled by a servo-mechanism in order to maintain constant the tip-speed-ratio of the windmill, and therefore an optimum efficiency.
The stirrer can be compared to a fluid-brake. Its temperature is increased by the friction of the internal fix and rotating intricated paddles, along hours and days of continuous operation.
The electronic circuit of the servo-mechanism is fed by the currents generated by an anemometer and a tachometer coupled on the shaft of the windmill.
Both signals are processed by an electronic integrator that filters and averages the quick instantaneous variations of the wind over about one minute, in order to avoid oscillations of the system.
The signal resulting from the combination of those signals is processed to make it representative of the instantaneous U/V tip-speed ratio of the windmill, and to make it linear.
The output is compared to a target voltage representative of the optimum tip-speed ratio, and a signal is issued to control the step-by-step motor of the water-gate, so that its opening angle is increased or decreased in view of cancelling the previous differential signal (positive or negative).
In the lack of wind, or for very low wind speeds, the intensity of the current generated by the anemometer remains below an adjustable threshold, which results in the complete opening of the water-gate and the emptying of the stirrer since there is no more ingress of water by the water-pump. Further starting of the windmill is therefore eased, since no resistive torque is applied to the shaft.
As soon as a sufficient signal appears at the output of the anemometer, the water-gate is completely shut-down and the stirrer is filled untill when the above-mentioned regulation starts again.
This curve shows how the servo-mechanism functions: by acting on the resistive torque applied to the shaft, the rotating-speed "N" of the windmill is continuously adjusted so that the operating point stays on the maximum power (dot line) corresponding to the actual wind-speed.
When the wind becomes steady, the exit of water by the water-gate (as controlled by the servo-mechanism) equals the ingress of water by the water-pump. The level of water in the stirrer remains constant.
The stirrer must be sized to contain enough water to create a sufficient resistive torque when the maximum wind is blowing. This wind-limit arises from the capacity of the tower to withstand the corresponding drag-moment on the windmill.
The reactivity of the servo-mechanism and its efficiency would have to be tested and may be improved on site: integration-time of the anemometer measurement , water ingress rate versus wind-speed, throughput and linearity of the exit water-gate ... In the preliminary design, we have assumed an efficiency of 80%.
2.2_Alternatives for the water-stirrer
An alternative to the voluminous water-stirrer described hereabove, could be to use an off-the-shelf water-brake. Some of them that are in the range of the power to be absorbed for the windmill optimum operation. They can be found in the field of car engines accessories.
The first one is a water-brake absorber, generally used as a dynamometer to measure the torque of car-engines on a test bench. The second one is a device named "retarder" which is normally used on the trucks to slow-them down on steady slopes, in order to save their normal disk-brakes life-time.
Both use a wheel with specific cavities rotating within a "stator" designed with matched cavities. Water (or sometimes oil) is inserted in the device and the friction of the liquid squeezed between the rotor and the stator's cavities generates the braking effect. The more the quantity of liquid introduced in the device, the more the braking effect. The rated absorbed power is obtained when the device is full of liquid. The liquid is heated in the process, so, in practice, a flow of liquid from a storage tank, has to be continuously circulated by a pump.
Both devices are compact enough to be mounted directly on the horizontal shaft of the windmill, on top of the tower. The adaptation to our project is straight-forward:
The resistive torque is easily controlled by adjusting the amount of water in the device. The windmill is thus maintained at its most efficient point of operation.
The resistive torque is easily controlled by adjusting the amount of water in the device. The windmill is thus maintained at its most efficient point of operation.
The water of the underground seasonal storage will be circulated in the brake, and thus benefit from the produced heat. An immersed water-pump is required and some flexible and insulated pipes need to be installed along the tower.
The main potential problem is that the rotating speed of the windmill is not in the range of the RPM of a car-engine, and some adaptation would probably have to be realized (introduction of a gear-box) and validated in real conditions.
1) An example of water-brake absorber:
An example of such a water-brake absorber is the DYNOmite 5":http://www.dynomitedynamometer.com/a...e_absorber.htm
2) An example of "retarder" normally used on the trucks as a secondary brake.
The manufacturer VOITH proposes two suitable models although somehow oversized for our project.
The Secundary Water Retarder SWR which works directly with the coolant water of the engine, and the R115E which uses oil for the process and then an heat-exchanger for the transfer to water.
The manufacturer VOITH proposes two suitable models although somehow oversized for our project.
The Secundary Water Retarder SWR which works directly with the coolant water of the engine, and the R115E which uses oil for the process and then an heat-exchanger for the transfer to water.
The control of the quantity of liquid admitted in the system to adjust the resistive torque, is pneumatic and requires a specific interface to be developed.
3) Eddy-current brakes have not been considered since they are electrically-driven, and generally air-cooled, which is not directly compatible with the purpose of this project.
3) Eddy-current brakes have not been considered since they are electrically-driven, and generally air-cooled, which is not directly compatible with the purpose of this project.
For comparisons with the characteristics of the above-mentioned water-brakes, here are some figures relative to the 5m diameter, multiblades, low-speed windmill used in this project:
5m/s 19RPM 600W/0.8HP 285N.m
10m/s 38RPM 4.6kW/6.2HP 1140N.m
17m/s 65RPM 22.5kW/30HP 3300N.m (the SWR is limited to 3500N.m max.)
25m/s 95RPM 71kW/95HP 7150N.m
The interest of such alternative devices is that they are standard and compact products, but they are rather sophisticated, expensive and bound to failure.
Conclusion: By comparison the proposed water stirrer is very voluminous and needs to be installed inside the building via a rear 4x4 axle and a long power-transmission shaft. But it is a low technology device, cheap and easy to manufacture in a local workshop.
2.3_ Comparison with a solution implementing a standard wind-generator and heating resistors immersed in the water-tank
There is an alternative solution for heating a building, that could be compared to the above one in terms of the optimization of the wind power recovered whatever the prevailing wind speed.
That solution would implement a high-speed wind turbine, associated with a synchronous generator interconnected to the grid through an electronic interface. The heating resistors are connected at the output of the inverter.
The wind-turbine is allowed to operate at variable speed, since the three-phase output of the generator is first rectified and then get through a power-inverter to interface with the grid. So the fix frequency of the grid is made independant from the rotational speed of the wind-generator.
Moreover, the back-to-back rectifier-inverter includes a MPPT controller (Maximum Power Point Tracking) that allows the wind turbine to operate permanently (whatever the wind-speed), at its optimum tip-speed ratio, and therefore produce its maximum power.
In fact, this MPPT controller adjusts in real time the resistive torque applied to the wind-turbine shaft by the generator, by enforcing the instant power to be transferred towards the grid (the current drawn on a generator creates a proportional resistive torque).
by acting on the rectifier's "time of conduction"(or PWM Pulse Width Modulation), the controller-1 enforces the optimum torque on the wind turbine shaft by determining the corresponding electrical load on the generator. For this purpose, it uses a reference map reflecting the curve of the successive maxima for every wind-speeds.
The controller-2 allows this amount of power to be transferred to the grid, by acting on the inverter's "time of conduction" (PWM).
Note: Hopefully, the grid can accept instantaneouly any amount of power (active and reactive), which would not be the case with a local distribution network.
Some manufacturers like Scirocco/Eoltec for instance propose such a complete package including a 5.6m 2-blade wind turbine with centrifugal pitch-control, a 6kW synchronous generator and the rectifier-inverter-MPPT interface with the grid.
By implementing this system we are ensured that the wind turbine transfers the maximum recoverable power to the grid at any wind-speed. But this is not exactly our purpose. What we want here is to use all this power to heat the building using resistors immersed in the water-tanks.
Note: Illustrations are from Intech editions ( Ali M. Eltamaly, A. I. Aloha and Hassan M. Farh )
An alternative and simpler solution, although not as efficien as the previous one, is to use a simple wind-turbine without all the attached electronics to interface the grid and without the attached MPPT to optimize the recovered power. The resistors are simply and directly connected at the output of the generator.In that case the produced power depends only of the ohmic value of the resistors and of the internal impedances of the generator. The windmill will rotate at the speed dictated by the resistive torque created by the current drained by the resistors. The value of the resistor must be chosen so that the rated power of the generator is not exceeded under strong winds.
A significant enhancement is obtained if the magnetic field created in the rotor can be adjusted (once for all), so as to match as much as possible with the maxima of power for every wind-speed.
The effectively recovered power is mentionned by the blue or red points on the curves.
Your solution looks very complicated and constraining, when using a standard wnd-generator with a simple electric resistor could be a lot more straight-forward to heat the water-storage.
Have you ever considered this solution ?
Your question is interesting and of course, I had it in mind continuously when I started developing my own project.
First of all I would like to mention that wind power (electrical or thermal) is definitively not the right solution for heating a house, when it is in competition with solar energy. It is the reason why I suggest to reserve it to nordic countries like Scotland, Denmark or other Scandinavian countries (Patagonia or Cape-Town in the Southern hemisphere) where sunshine periods are scarce in winter (when the most you need heating !).
A second point is that if you decide to use wind power to feed a resistor to heat a water-tank, the main problem will be to keep the wind-generator rotating at the optimum speed. If it is not the case, the yield of the generator will be optimum only a few times per day, and the overall efficiency very poor. The solution would be either to control permanently the tilt of the blades (which is not very common in that range of wind-generators), or to use a variable resistor controlled by a servo-mechanism that you will have to invent.
Third point is that a wind-machine must have a high speed of rotation to drive the generator, which means a sophisticated profile for the blades and therefore a higher cost of production. The generator itself is also costly and bound to failures.
A fourth but maybe marginal point, is that a wind-turbine is more noisy than a windmill due to its higher rotational speed. It could be a nuisance if it is installed too close to the house to be heated.
Fifth and last point, the heat-losses dissipated by the electric generator cannot be recuperated in the storage, while with the windmill/stirrer solution 100% of the available power is recuperated.
In the solution I propose, all the components are easy to produce in a local workshop, and reliable. Of course the prototype needs some more efforts to be fully developed, but once it is validated, it will be very easy and cheap to reproduce.
It is true that the building is very specific and probably rather expensive to construct. It also requires to reserve the whole basement to the water-tanks (although they could maybe advantageously be turned into a covered leisure-pool for recreational use).
Anyhow, I think that this solution is more directed to public or professional premises, like for instance a place where industrials in the field of renewable energies, could have an office and exhibition-hall for their products in live situation. The proposed building is rather spectacular and may attract many visitors in the region, which is coherent with a commercial use.
I have a question relative to the fluid you use in your system. Why not using lubricating oïl in your system instead of water, at least in the stirrer ? I think you would have a better efficiency.
The windmill transfers to the stirrer a certain amount of mechanical work, which is equivalent (according to the physicist James Prescott Joule in 1843) to a certain quantity of heat.
This quantity of heat will result into an elevation of the temperature of the liquid that is used (oil or water). For a same volume of liquid, the different specific heats will result into a higher increase of the oil temperature, compared to water. But, symmetrically, the decrease of the temperature when heating the building, will be quicker with oil than with water.
At the end there is no miracles ! Anyhow the overall quantity of heat transferred to the building will be the same in both cases.
As a reminder, the specific heat is 4180 J/kg/°C for water, and 2000 J/kg/°C for oil.
Since the comparison has to be done for identical volumes and not for identical masses, the density of the liquids has also to be considered. Water density is 1kg/L, oil density is 0.85 kg/L.
It results that, for a given quantity of heat transferred to the system, the maximum elevation of temperature will be 2.45 times higher with oil than with water.
But this constitutes a drawback rather than an advantage, and the first conclusion is that the use of oil instead of water in the complete system would result into the necessity to improve the insulation of the walls of the storage-tanks to minimize the heat-losses (that are proportional to the delta-T versus ambient temperature).
Anyhow, several other obvious disadvantages to store such large volumes of oil, would have precluded its use in a manned building: the risk of fire in case of a flash of lighting striking the metallic tower and windmill, the risks of pollution, the smell, the negative image of fossil energy for a project considered to be ecological, the overall cost of oil, the eventual formation of emulsions and foam when stirring energetically the oil during long periods, and of course the regulations in force for inhabited places.
Nevertheless, the question is still open to the opportunity to reserve the use of oil to the stirrer without extending it to the whole heat-storage.
Indeed, there are some advantages for the use of oil within the stirrer: lubricating of the gears and ball-bearings, natural protection against rust and corrosion of the components, better efficiency of the friction between fix and rotating paddles due to its viscosity.
Nevertheless this alternative would require to implement a completely sealed and proof stirrer and heat-exchanger between oil and the water of the storage. This is not easy to realize and would result to be much more expensive, for only a few advantages.
By comparison, the water-stirrer does not require to be perfectly waterproof, since anyhow the water-gate permanently flushes water out of the stirrer to the down-below main reservoir. Eventual leaks of the stirrer will not affect the operation at all.