Wind Turbine Hydraulic to Electrical Conversion

[Jon Richfield]

... youtube video where he describes the vane version, it begins at minute 35:45

He describes the concept in many of the recent articles that he has posted on his website: www.danhelgerson.com , particularly the one titled, "Transformation Complete".

Thanks, Interesting video. Still going through the whole one. Good luck if the info I regurgitated is of value.

 

[deckart]

Thanks, Interesting video. Still going through the whole one. Good luck if the info I regurgitated is of value.

It's interesting to see that the idea has been around a long time. It's a very simple concept that is common in electronics but just hasn't been exploited in fluid power.

 

[Baluncore]

it sounds like the VDT (Variable Displacement Transformer) concept that a friend and I are working on.

In post #24, I suggested a constant pressure, variable volume pump driven by the wind turbine, with the fluid driving a constant speed hydraulic motor. That is really a VD transformer with a hydraulic loop connecting two rotating shafts.

If instead you have a fluid input loop, couple the shafts together and have a fluid output loop, then it is more obviously a VDT described from the viewpoint of an hydraulic engineer.

In the wind generator application there is a shaft input and a shaft output, coupled by a pump and motor hydraulic circuit. Inserting a VDT into that loop would be less efficient than migrating the transformation to the ends of the system by using a VD CP pump and a VD fixed speed motor. That technology has been available for many decades. These days there are more efficient electrical solutions.

 

[deckart]

In post #24, I suggested a constant pressure, variable volume pump driven by the wind turbine, with the fluid driving a constant speed hydraulic motor. That is really a VD transformer with a hydraulic loop connecting two rotating shafts.

If instead you have a fluid input loop, couple the shafts together and have a fluid output loop, then it is more obviously a VDT described from the viewpoint of an hydraulic engineer.

In the wind generator application there is a shaft input and a shaft output, coupled by a pump and motor hydraulic circuit. Inserting a VDT into that loop would be less efficient than migrating the transformation to the ends of the system by using a VD CP pump and a VD fixed speed motor. That technology has been available for many decades. These days there are more efficient electrical solutions.

The reason this has not been done is that the input velocity is too low for a variable displacement constant pressure pump to operate. It is not a feasible circuit.

Pumps are not designed to operate at such low velocities which is why I designed it using hydraulic cylinders (0-100 RPM maybe, the lower the better). What happens on the other end, at the generator, has to be done at higher velocities that are typical for generating electricity.

The electrical problem is dealing with the low and inconsistent wind speeds. It is not that efficient, from what I've read. Using hydraulics, you capture more of the energy at any speed. As long as the there is motion, there is energy being converted to hydraulic energy. And that is easily converted back to mechanical energy to drive a generator.

So, I disagree that the current electrical solutions are more efficient or less expensive. At least to the point that it is worth research. I have not seen any studies regarding this. My opinion is that because the product is electrical energy that the development of these types of devices has been electrical-centric and the use of fluid power as a means of interim power transmission has not been explored.

What you described in post #24 is conceptually the same thing it's just not using the same components. Hydraulic pumps, in general, are designed to be driven at velocities typical of fuel or electrically driven prime movers. 1200-3600 RPM.

 

[Jon Richfield]

It's interesting to see that the idea has been around a long time. It's a very simple concept that is common in electronics but just hasn't been exploited in fluid power.

Yes, I can't remember whether James showed it to me in 1968 or 1969, but I think he had had the idea some years before. I sometimes wonder whether something of the kind shouldn't be more efficient and versatile than our current transmission in automatic cars.

 

[Baluncore]

My opinion is that because the product is electrical energy that the development of these types of devices has been electrical-centric and the use of fluid power as a means of interim power transmission has not been explored.

I remember reading reports in the 1970s of prototypes of hydraulic systems using VDT being built and evaluated in the USA and UK for wind generators. Even with hydraulic transformers, they were not as efficient as multi-pole alternators and I believe were upgraded or sold. That is probably why you have not come across them.
There were a couple of bright engineers in my lab who evaluated hydraulic VDTs to improve the efficiency of wind turbine generators, but they could not show a competitive advantage over existing electrical alternators.
The subject keeps coming up as a search of google images for 'hydraulic wind turbines' will show you.

 

[anorlunda]

The primary advantage of using hydraulics is going to be how it deals with a varying input RPM and load, and the energy conversion is very direct.

When you talk about advantages of one design versus another, starting with RPM as the independent variable masks part of the problem. You should start with available energy. At low wind speeds, low energies make power production unattractive regardless of the efficiency of the mechanical/hydraulic/electrical mechanisms.

Instead, the focus should be on efficiency in the attractive range of wind speeds, roughly 15-25 mph (7-11 mps). So, the hydraulic method might still be interesting, but not at low wind speeds. Also, there are freedoms in the design of the wind turbine (such as variable pitch) as well as the conversion mechanisms, so you have to evaluate the entire system to determine attractiveness.

Conservation of mass requires that the amount of air entering and exiting a turbine must be equal. Accordingly, Betz's law gives the maximal achievable extraction of wind power by a wind turbine as 16/27 (59.3%) of the total kinetic energy of the air flowing through the turbine.[15]

The maximum theoretical power output of a wind machine is thus 16/27 times the kinetic energy of the air passing through the effective disk area of the machine. If the effective area of the disk is A, and the wind velocity v, the maximum theoretical power output P is:

{\displaystyle P={\frac {16}{27}}{\frac {1}{2}}\rho v^{3}A={\frac {8}{27}}\rho v^{3}A}

,
where ρ is the air density.

 

[deckart]

I remember reading reports in the 1970s of prototypes of hydraulic systems using VDT being built and evaluated in the USA and UK for wind generators. Even with hydraulic transformers, they were not as efficient as multi-pole alternators and I believe were upgraded or sold. That is probably why you have not come across them.
There were a couple of bright engineers in my lab who evaluated hydraulic VDTs to improve the efficiency of wind turbine generators, but they could not show a competitive advantage over existing electrical alternators.
The subject keeps coming up as a search of google images for 'hydraulic wind turbines' will show you.

In the 70's variable displacement pumps and motors were not common at all. Most hydraulic applications used low efficiency fixed displacement devices. One of the first articles that came up when I googled "hydraulic wind turbine" is this: http://www.machinedesign.com/energy/hydraulic-wind-turbines describing very similar configuration but the author is considering the use of traditional high RPM pumps which are inefficient to drive at low speeds. I address this by simply using hydraulic cylinders which are 99% efficient.

You have not convinced me that this approach has been evaluated. Also, the VDT is a theoretical device that has never been commercially manufactured. I believe you are thinking of a hydrostatic transmission. The variable pump/motor relationship is virtually the same concept.

 

[deckart]

When you talk about advantages of one design versus another, starting with RPM as the independent variable masks part of the problem. You should start with available energy. At low wind speeds, low energies make power production unattractive regardless of the efficiency of the mechanical/hydraulic/electrical mechanisms.

Instead, the focus should be on efficiency in the attractive range of wind speeds, roughly 15-25 mph (7-11 mps). So, the hydraulic method might still be interesting, but not at low wind speeds. Also, there are freedoms in the design of the wind turbine (such as variable pitch) as well as the conversion mechanisms, so you have to evaluate the entire system to determine attractiveness.

The hydraulic method IS interesting because of its efficiency at low wind speeds (using a hydraulic cylinder pump) vs electrical-mechanical. Input energy is a different conversation. I'm proposing a different method of transmission.

How hydraulic energy is converted back to mechanical energy to drive a generator can be done with a common variable displacement motor as opposed to a VDT. In fact, I should change that in the schematic to prevent any confusion around a device that doesn't even exist yet. The advantage of a VDT is that it can be used to maintain a target RPM passively. With currently available components the RPM would be controlled by modulating the displacement of the motor in a closed-loop.

 

[deckart]

For various personal (abstract, not commercial nor professional) reasons I find this discussion very interesting
(see i.a. http://fullduplexjonrichfield.blogspot.co.za/2017/04/heavier-duty-banking-appendix-supplement.html)
And I called to mind a discussion about 48 or 50 years ago, when a then colleague in IBM, one James Philbrick (very intelligently creative, but since deceased, I am very sorry to say) described an invention that I admired. I know that he patented it in the same or following year, but I never saw that anyone took it up. I mention it here, as nearly as I remember it, in case anyone can put it to constructive use. Alternatively, if anyone happens to know that it is currently in use, I would be curious to know. Otherwise it would be a pity to waste it, I reckon.

James had been a hydraulics engineer with experience in designing systems for ships, and had recognised that a generalisation of the rotary vane pump could be used to power one flow of liquid by the input of another flow. By shifting the axis of rotation of the rotor, one could smoothly in effect change gear, either moving more fluid against lower resistance, or less fluid against higher resistance. I include a sketch of what I can remember from my distant youth. I omit the notional mechanism for moving the axis of the rotor (James represented it at the time simply as an external lever of type two, with the rotor axis in the middle).

The power is applied by one pair of opposed inlet-outlet channels (say the green arrows) which drives fluid through the other pair according to the position of the rotor in the outer drum. No doubt an arbitrary set of input-output channels could be combined for more complex requirements.

For your attention for what it is worth.

I apologise, but I failed to insert the diagram in usable form. In case my description fails to convey anything articulate (very likely!) I have posted it at:

https://www.facebook.com/photo.php?...5908628.111317.100001576411737&type=3&theater

Good luck to anyone interested.

If anyone could tell me how to include the image more conveniently, feel welcome.

Jon, can you reference the patent number? I am not finding it.

 

[Jon Richfield]

Jon, can you reference the patent number? I am not finding it.

I have not the faintest clue, sorry. I never knew that. James just mentioned it to me afterwards. I seem to remember it was a world patent or something, and we were in South Africa. I know nothing about the practicalities of patents now, and less than that then. It would have been about 1968 or 1970.

 

[Baluncore]

but the author is considering the use of traditional high RPM pumps which are inefficient to drive at low speeds.

Obviously high speed pumps will be optimised for high speeds. The same type of pump can be scaled to operate at low speeds. A bent-axis swash-plate motor uses hydraulic cylinders with variable stroke. The swash plate was and still is most commonly used for adjustable rate pumps and motors. There is little difference between axial hydraulic cylinders with a swash-plate and your proposed radial configuration.

... I address this by simply using hydraulic cylinders which are 99% efficient.

The movement of hydraulic fluid through the lines and valves connected to the hydraulic cylinder is where the majority of the loss occurs. Those ancillary losses usually amount to between 10% and 20%. Long hydraulic lines between pump and motor will greatly improve oil cooling and oil life, but will also reduce the efficiency overall.

Hydraulic systems have a very high power to weight ratio, but that comes with poor efficiency. When you double the RPM of a hydraulic pump or motor you double the power for the same weight, while the flow is doubled at the same pressure. That is why small high speed motors are more common than big, heavy and very expensive radial cylinder pumps or motors like your design. Since hydraulic losses increase in proportional to the square of the fluid velocity. Bigger, more expensive hydraulic lines and valves must be used not only with big and slow rotary devices, but also with small and fast rotary systems. The maximum pressure is also limited for larger diameter hoses. That cuts the maximum power transmitted.

You have not convinced me that this approach has been evaluated. Also, the VDT is a theoretical device that has never been commercially manufactured. I believe you are thinking of a hydrostatic transmission. The variable pump/motor relationship is virtually the same concept.

What do you mean by “hydrostatic transmission”, a toroidal flow torque converter? I see any hydraulic pressure transmission as a “hydrostatic transmission”, be it a VR pump and VR motor combination or the return of the VDT concept under another name.
“The variable pump/motor relationship is virtually the same concept” as what? The newly named VDT or the combination of VR pump and VR motor I suggested. I think they are the same. Your requirement was to have hydraulic flow up and down the tower. That requires a separate pump and motor, with a fluid reservoir = header tank and a filter system high on the tower where fluid must be pushed into the pump inlet.

One of the reasons I'm sharing this is because I believe fluid power systems are being overlooked in the renewable energy industry.

This has been a common theme. Every five years there is someone who jumps on the hydraulics bandwagon. They focus on rediscovering hydraulic technology and pump design. Then they go quiet when they find out how inefficient hydraulic transmission systems can be. A hydraulic system is not as efficient, nor as flexible as today's electrical technology, control systems and accumulator technology.

 

[deckart]

Obviously high-speed pumps will be optimised for high speeds. The same type of pump can be scaled to operate at low speeds.

Sure, high-speed pumps can be scaled to operate at low speeds. But they aren't. There is no market for them and that is why they don't exist. Pumps require a prime mover, something to drive them. The common items that drive pumps operate at high-speeds, i.e. gas engines, electric motors. There are no prime movers to drive low-speed pumps. What would be the point? Where is the market? So, I propose a pump design that CAN operate efficiently at low speeds. It is just a combination of common, inexpensive, hydraulic cylinders. I didn't redesign a wheel. I just made a low-speed pump suitable for this application.

A bent-axis swash-plate motor uses hydraulic cylinders with variable stroke. The swash plate was and still is most commonly used for adjustable rate pumps and motors. There is little difference between axial hydraulic cylinders with a swash-plate and your proposed radial configuration.

Actually, there are significant differences. One of which is that I'm not adjusting flow (or in your words, "rate"). This is a fixed displacement device. Flow is dependant on RPM, RPM is the variable. The problem this solves is that it captures energy efficiently regardless of how low the RPM is. Electro-mechanical systems do not do this efficiently as you claim. The systems that are used employ a lot of expensive techniques to address this. Some of them are very ingenious from what a colleague has described to me, though I can barely follow the theory behind it.

The movement of hydraulic fluid through the lines and valves connected to the hydraulic cylinder is where the majority of the loss occurs. Those ancillary losses usually amount to between 10% and 20%. Long hydraulic lines between pump and motor will greatly improve oil cooling and oil life, but will also reduce the efficiency overall.

This is a very ambiguous paragraph. There are always losses when you transmit energy, whether electrical, mechanical, or hydraulic. Let’s say there is a total of 10% line and valve loss. Which is high, imo, because the only valves I’m using for the work lines are check valves. Consequently, I may be capturing 30% more energy. I really don’t know yet but I have a system that could be used to find out.

Hydraulic systems have a very high power to weight ratio, but that comes with poor efficiency. When you double the RPM of a hydraulic pump or motor you double the power for the same weight, while the flow is doubled at the same pressure. That is why small high speed motors are more common than big, heavy and very expensive radial cylinder pumps or motors like your design. Since hydraulic losses increase in proportional to the square of the fluid velocity. Bigger, more expensive hydraulic lines and valves must be used not only with big and slow rotary devices, but also with small and fast rotary systems. The maximum pressure is also limited for larger diameter hoses. That cuts the maximum power transmitted.

This is riddled with ambiguity and incorrect generalization. With higher pressures densities you transmit the same power with less flow, smaller lines. Just as you can deliver the same power with a higher voltage and less current, smaller lines. Regardless, extremely large ID high pressure hose is available (see Parker 797 series 6000 psi hydraulic hose).

What do you mean by “hydrostatic transmission”, a toroidal flow torque converter? I see any hydraulic pressure transmission as a “hydrostatic transmission”, be it a VR pump and VR motor combination or the return of the VDT concept under another name. “The variable pump/motor relationship is virtually the same concept” as what? The newly named VDT or the combination of VR pump and VR motor I suggested. I think they are the same.

Fine, semantics. It isn’t really the point of the system.

Your requirement was to have hydraulic flow up and down the tower. That requires a separate pump and motor, with a fluid reservoir = header tank and a filter system high on the tower where fluid must be pushed into the pump inlet.

Up/down, yes. No, it does not require a separate pump and motor. I’m using the radial pump to circulate all fluid in the system. Filtration, yes, and yes, the reservoir will be pressurized to aid pump inlet delivery. It is a closed system. Much like a hydrostatic transmission, in fact. And remember, fluid head coming down the tower is the same as is going up, energy-wise, they are equal.

This has been a common theme. Every five years there is someone who jumps on the hydraulics bandwagon. They focus on rediscovering hydraulic technology and pump design. Then they go quiet when they find out how inefficient hydraulic transmission systems can be. A hydraulic system is not as efficient, nor as flexible as today's electrical technology, control systems and accumulator technology.

You still haven’t proven your point beyond making vague generalizations.

 

[deckart]

This forum makes for a good chalkboard!

I've been doing some more research into the challenges of current large-scale wind turbine arrays. One of the problems, of course, is energy storage. There are often demands when the wind isn't blowing. There are meteorological projections considered in advance for opportune times to bring the arrays online. Tesla and their mega-battery systems, as demonstrated in Australia, work great in this capacity.

The wind-turbines themselves can provide a good piece of energy storage when a hydraulic system is utilized with the use of the basic hydraulic accumulator. A good deal of energy can be stored for use during those opportune periods. The limitations are only in the capacity of accumulators that are utilized and with the large numbers wind turbines that some of these wind farms employ, this can be a very substantial amount of energy.

Here is a schematic illustrating how such a system might be configured:

 

[NTL2009]

What would these accumulators physically look like? I assume these are oil pumped into a chamber with air/gas, so the energy storage mechanism is compressed gas?

I was under the impression there is a fair amount of loss in a system like that. Compressing a gas heats it, and that heat is lost over time, which could be many hours if you are trying to smooth wind differences over the course of the day.

How large would accumulators be per MW-Hr of storage?

 

[Baluncore]

deckart; You show accumulators that have a reciprocal pressure:volume relationship. The compression of a gas in the accumulator will lead to thermodynamic inefficiency. The changing pressure, while operating over the energy storage range, will make it more difficult to optimise the efficiency of the hydraulic pumps and the generator motor.

When a system is operated at a constant pressure, it can use a pumped water reservoir with a reasonably stable operating pressure due to the fixed head. There is a parallel here with the fixed voltage of the electricity grid and distribution system.

One requirement of an accumulator is the need for a greater working volume of fluid, plus a companion reservoir with an equal fluid capacity. If water is the low cost environmentally friendly fluid employed for the storage of potential energy, then there needs to be a reservoir low pressure "tank" at the altitude of the wind farm pumps, with a significantly higher altitude “accumulator” lake. Those fluid storage lakes should preferably have very large surface areas so as to minimise pressure variation with energy stored, and to minimise disturbance to the natural environment.

An underground water reservoir, or a fabric bag held deep underwater, will have the same hydrostatic pressure change in the connection line to the deep storage as the storage itself. That will make energy storage impossible unless a lower density coupling fluid such as compressed air is used. That comes with the thermodynamic inefficiency of the gas compression and expansion cycles.

 

[deckart]

What would these accumulators physically look like? I assume these are oil pumped into a chamber with air/gas, so the energy storage mechanism is compressed gas?

I was under the impression there is a fair amount of loss in a system like that. Compressing a gas heats it, and that heat is lost over time, which could be many hours if you are trying to smooth wind differences over the course of the day.

How large would accumulators be per MW-Hr of storage?

Here is an example of what an accumulator stand might look like for a large wind-turbine station:

Thermal losses are determined by the rate of charge and discharge. A system sized for slow charge and discharge will mitigate the impact of thermal losses. Inefficiency will depend on how the accumulators are used. This is the difference between an isothermal and adiabatic application of an accumulator.

(Actually, I'm wrong on this. Heat is generated regardless of rate. A solution, the accumulator cylinders need to be insulated so that the heat energy is kept in the system and the accumulator pre-charge and size would have to reflect the heat-induced gas expansion.)

"The primary cause of capacitance in a hydropneumatic accumulator is compressibility of the gas. The behavior of the gas can be approximated with the ideal gas laws under isothermal (constant temperature) or adiabatic (no heat energy enters or leaves the system) conditions." Source: http://www.hydraulicspneumatics.com...tors/Article/False/7225/TechZone-Accumulators -Jack L. Johnson, P. E.

Considering we are storing energy that would otherwise not be generated by the wind-turbine and stored at all, this is a great application for energy storage for use later in a more scheduled fashion.

I will crunch some numbers on what the MW-Hr storage capacity could look like.

 

[deckart]

deckart; You show accumulators that have a reciprocal pressure:volume relationship. The compression of a gas in the accumulator will lead to thermodynamic inefficiency. The changing pressure, while operating over the energy storage range, will make it more difficult to optimise the efficiency of the hydraulic pumps and the generator motor.

When a system is operated at a constant pressure, it can use a pumped water reservoir with a reasonably stable operating pressure due to the fixed head. There is a parallel here with the fixed voltage of the electricity grid and distribution system.

One requirement of an accumulator is the need for a greater working volume of fluid, plus a companion reservoir with an equal fluid capacity. If water is the low cost environmentally friendly fluid employed for the storage of potential energy, then there needs to be a reservoir low pressure "tank" at the altitude of the wind farm pumps, with a significantly higher altitude “accumulator” lake. Those fluid storage lakes should preferably have very large surface areas so as to minimise pressure variation with energy stored, and to minimise disturbance to the natural environment.

An underground water reservoir, or a fabric bag held deep underwater, will have the same hydrostatic pressure change in the connection line to the deep storage as the storage itself. That will make energy storage impossible unless a lower density coupling fluid such as compressed air is used. That comes with the thermodynamic inefficiency of the gas compression and expansion cycles.

Yep, these are good ideas and would work well in areas that have large volumes of water available and the geography to support storage.

With hydraulic accumulators, the energy density can be very much higher. 100 psi of head pressure would require 50 times the volume of fluid that is can otherwise be stored at 5000 psi, for example.

 

[deckart]

I estimate that the accumulator bank illustration with the man standing in it is approximately 1000 gallons.

I contacted a friend who has had to make these calculations before in his consulting work. This is what he sent me:

One complete charge would power the average US home for more than 3 years. This seems considerable but I'm not sure what percentage of a towers capacity factor (avg is 25% of rating) would be useful for day-to-day use. The "capacity factor" is the average output relative to a wind-turbines rating.

 

[NTL2009]

I estimate that the accumulator bank illustration with the man standing in it is approximately 1000 gallons. ...

[ chart shows 36.25 kW-hr]

One complete charge would power the average US home for more than 3 years. ...

Please check your (and my) math.

Google says:

In 2016, the average annual electricity consumption for a U.S. residential utility customer was 10,766 kilowatthours (kWh), an average of 897 kWh per month.

That is ~ 30 kW-hr per day, so that 1000 gallons would power the average US home for a little more than 3 years one day ( ~ 28.8 hours).

edit/add: I realize now you are likely off by a factor of 1000 (KW versus MW or W), as three years is ~ 1000 days.

 

[NTL2009]

If my math was right above, let's put that in perspective with some round numbers (going from memory here, should be close enough) for illustration:

I think a typical Wind Turbine is rated ~ 1 MW (close enough?), the capacity factor might be ~ 30% (close enough?), so on average it is providing 300 KW. If we also round the US home average down to 1 KW average (that would be 24 k-Wh per day), that means a single turbine is supplying energy for ~ 300 homes. So a single turbine would require 300 of those big units the man is standing next to in order to power the homes it supplies for ~ 1 day.

What does one of those accumulator arrays cost?

 

[NTL2009]

Taking that a step further, if the accumulator captured 36.25 k-Wh of excess ('free') electricity every single day, and was able to sell it during peaks every single day at an average wholesale price of 10 cents / k-Wh, that would be $3.625 in sales each day (ignoring losses since I'm rounding/guessing at rough numbers anyhow). That would be ~ $1,325 per year.

I know nothing about the cost of an accumulator like that, but my gut tells me it is far too large of a capital investment (let alone maintenance costs) to justify a $1,325 annual income. I'm quite certain it would be negative after factoring in the cost of capital. Even if you could float a bond at 3%, you don't hit break-even (not even factoring maintenance costs) unless you keep the cost below $44,000.

All these storage methods sound enticing, until we do the math.

 

[Baluncore]

All these storage methods sound enticing, until we do the math.

Then consider the possibility of pumped underground energy storage below a wind farm.

Find a raised area suitable for a wind farm, over an aquifer, with the water table at say 100 metres below. Take advantage of the water table being set by springs at the foot of the slope. Bore a hole down to the water table, plus an allowance for draw-down in dry seasons. To store energy, pump water from the aquifer below, up into a wide area reservoir on the surface. To recover the stored energy, return water from the reservoir to the aquifer below, so driving the pump.

The water transfer lift pump will need to be at the bottom of the hole where it will not cavitate. The pump will always operate with a reasonably stable head, when as a pump, or when as a motor. Drive to the pump could be mechanical, electrical or hydraulic. It could use hydraulic oil in a closed cycle, in parallel with the wind turbine and electric generator. The hydraulic oil pressures will be significantly higher than the pumped storage hydrostatic pressure. The ratio being built into the drive to the down-hole pump. That integrates well with a constant pressure hydraulic fluid system.

One problem with hydraulic drive is that hydrostatic pressure rises with depth which requires more expensive higher pressure rated pipes be used at greater depths in the dry part of the bore hole. If that is a problem then the hydraulic drive could be through coaxial pipes with the HP drive inside the LP drive fluid return line, all inside the outer water transfer pipe. The advantage of coaxial tubes is the reduction in maximum differential wall pressure. Only the outer transfer pipe may needs a heavier wall at depth.

Now to size the infrastructure for a 1 MW storage with a head of 100m.
Hydrostatic pressure = height * gravity * density. 100 m * 9.8 m/s² * 1000 kg/m³ = 980 kPa.
At 100m depth, the working pressure will be 980 kPa = 142 psi.
Design for an energy storage of 1 MW∙hr = 3600 MJ.
So it needs a reservoir of 3600 MJ / 980 kPa = 3.67 thousand cubic metres.
Make the reservoir a 31 m square, with an average depth of 4 metre = 3844 m³.
Have I got that about right, or did I slip 3 or 6 digits somewhere ?

It seems possible, but doing the math on the inefficiency of all the fluid flows may show it wastes significant energy.

There are places where artesian pressure prevents the above scenario. Some of those in remote Australia generate electrical power from the continuous high pressure flow of hot ground water, that then goes on to water livestock.

 

[deckart]

Please check your (and my) math.

Google says:That is ~ 30 kW-hr per day, so that 1000 gallons would power the average US home for a little more than 3 years one day ( ~ 28.8 hours).

edit/add: I realize now you are likely off by a factor of 1000 (KW versus MW or W), as three years is ~ 1000 days.

Got it! I was just looking at Watts. This wouldn't work in that capacity at all!

 

[NTL2009]

Then consider the possibility of pumped underground energy storage below a wind farm.
...
Now to size the infrastructure for a 1 MW storage with a head of 100m.
Hydrostatic pressure = height * gravity * density. 100 m * 9.8 m/s² * 1000 kg/m³ = 980 kPa.
At 100m depth, the working pressure will be 980 kPa = 142 psi.
Design for an energy storage of 1 MW∙hr = 3600 MJ.
So it needs a reservoir of 3600 MJ / 980 kPa = 3.67 thousand cubic metres.
Make the reservoir a 31 m square, with an average depth of 4 metre = 3844 m³.
Have I got that about right, or did I slip 3 or 6 digits somewhere ?

It seems possible, but doing the math on the inefficiency of all the fluid flows may show it wastes significant energy.

There are places where artesian pressure prevents the above scenario. Some of those in remote Australia generate electrical power from the continuous high pressure flow of hot ground water, that then goes on to water livestock.

I didn't check your math, but it seems right offhand. Pumped hydro is a much more practical approach. The high pressure accumulators seem more applicable to where space is constrained, or for a mobility/portability.

Got it! I was just looking at Watts. This wouldn't work in that capacity at all!

Well I wish it was right! :)

I was thinking it would then only take a 1/2 gallon system to store 12 hours of average household kW-hr. Someone on a TOD metering system could buy/store the cheap power all night, and then use it all day, and pay only the cheap rate (but only with net metering, as an inexpensive system could not handle peak loads, so it would need to run the meter backwards when the load was less than average, to make up for peak draw). That might work with a 1/2 gallon system, not so much with a 500 gallon system. I'm curious what a ~ 1 KW continuous duty motor/generator/pump would cost? Hmmm, I guess the pump/motor/generator is a fixed cost, only the tank size would change. Space constraints aside, I suppose a 500 G high pressure tank is far more expensive than a 1/2 gallon tank - that's lots of added area and force.

 

[deckart]

Well I wish it was right! :)

I was thinking it would then only take a 1/2 gallon system to store 12 hours of average household kW-hr. Someone on a TOD metering system could buy/store the cheap power all night, and then use it all day, and pay only the cheap rate (but only with net metering, as an inexpensive system could not handle peak loads, so it would need to run the meter backwards when the load was less than average, to make up for peak draw). That might work with a 1/2 gallon system, not so much with a 500 gallon system. I'm curious what a ~ 1 KW continuous duty motor/generator/pump would cost? Hmmm, I guess the pump/motor/generator is a fixed cost, only the tank size would change. Space constraints aside, I suppose a 500 G high pressure tank is far more expensive than a 1/2 gallon tank - that's lots of added area and force.

A hydraulic accumulator used for energy storage is simply a hydraulic cylinder without a cylinder rod. A floating piston separates the gas and hydraulic fluid. Here is a chart showing standard sizes from a supplier that I use, https://www.accumulators.com:

As far as cost for a small unit, you have the accumulator, a standard generator of an appropriate size from Lowe's or Home Depot, and an appropriate hydraulic motor and valve circuit. Remove the gas motor and install the hydraulics. I'll put together a BOM and get quotes together. Originally I was looking something small scale like this but then I saw how well this can scale up and spent my time on that. I believe it gets more cost effective at larger scales.

 

[cjl]

I think a typical Wind Turbine is rated ~ 1 MW (close enough?), the capacity factor might be ~ 30% (close enough?), so on average it is providing 300 KW. If we also round the US home average down to 1 KW average (that would be 24 k-Wh per day), that means a single turbine is supplying energy for ~ 300 homes. So a single turbine would require 300 of those big units the man is standing next to in order to power the homes it supplies for ~ 1 day.

What does one of those accumulator arrays cost?

It's even a bit worse than that - most newer utility-scale wind turbines are more like 2.5MW, and have capacity factors of 35% or so, so you're looking at the better part of a megawatt of continuous power. Really large offshore machines are fast approaching 10MW, and capacity factors of 50% are not unheard of, if you want to look at the extreme upper end of the market.

 

[cjl]

Actually, there are significant differences. One of which is that I'm not adjusting flow (or in your words, "rate"). This is a fixed displacement device. Flow is dependant on RPM, RPM is the variable. The problem this solves is that it captures energy efficiently regardless of how low the RPM is. Electro-mechanical systems do not do this efficiently as you claim. The systems that are used employ a lot of expensive techniques to address this. Some of them are very ingenious from what a colleague has described to me, though I can barely follow the theory behind it.This is a very ambiguous paragraph. There are always losses when you transmit energy, whether electrical, mechanical, or hydraulic. Let’s say there is a total of 10% line and valve loss. Which is high, imo, because the only valves I’m using for the work lines are check valves. Consequently, I may be capturing 30% more energy. I really don’t know yet but I have a system that could be used to find out.

There's nothing particularly hard to understand about modern wind turbine electronics. They fall into a couple of major categories, but they're fairly straightforward. Most wind turbines you see are geared, although a few are direct drive. The geared machines use a ~120:1 ratio gearbox to increase the shaft speed to somewhere in the neighborhood of 1800 RPM at full power. This is the speed at which the generator operates. Direct drive machines obviously have no gearbox, so they have very large diameter, multipole permanent magnet generators designed to operate at ~10-12 RPM at full power. In either case, full power efficiency is usually in the neighborhood of 90%. At low wind speeds, the rotors spin as low as ~4RPM, so the generators need to cover a speed range of about a factor of 3. This is well within the capability of modern generator design, and over the majority of the speed range, the generator efficiency is pretty similar to full power, around 90% (it's usually a bit more efficient at less than full power, but the details aren't important here).

There are also some additional losses associated with converting the electricity to the proper voltage and frequency for the grid, but that's also a very efficient process, whether you use full conversion (all the power goes from AC to DC and back to AC again), or whether you use a DFIG (https://en.wikipedia.org/wiki/Doubly-fed_electric_machine#Double_fed_induction_generator).

Now, it's true that below 10% of rated power or so, the efficiency drops, but the amount of additional energy you could gain is really just not worth it. Say you change the turbine from being 66% efficient at 5% of rated power to 90%. You've only increased power output by about 2% of the overall turbine rating, which will have a fairly small impact on annual energy production. I also suspect you would have a hard time with the hydraulic drivetrain concept in getting the full power efficiency up to 90%, so you'll probably lose more energy during high power production than you'll ever gain back in low wind, especially given that you would never want to put a wind turbine on a site where it spends most of its time at <10% power anyways.

 

[deckart]

There's nothing particularly hard to understand about modern wind turbine electronics. They fall into a couple of major categories, but they're fairly straightforward. Most wind turbines you see are geared, although a few are direct drive. The geared machines use a ~120:1 ratio gearbox to increase the shaft speed to somewhere in the neighborhood of 1800 RPM at full power. This is the speed at which the generator operates. Direct drive machines obviously have no gearbox, so they have very large diameter, multipole permanent magnet generators designed to operate at ~10-12 RPM at full power. In either case, full power efficiency is usually in the neighborhood of 90%. At low wind speeds, the rotors spin as low as ~4RPM, so the generators need to cover a speed range of about a factor of 3. This is well within the capability of modern generator design, and over the majority of the speed range, the generator efficiency is pretty similar to full power, around 90% (it's usually a bit more efficient at less than full power, but the details aren't important here).

There are also some additional losses associated with converting the electricity to the proper voltage and frequency for the grid, but that's also a very efficient process, whether you use full conversion (all the power goes from AC to DC and back to AC again), or whether you use a DFIG (https://en.wikipedia.org/wiki/Doubly-fed_electric_machine#Double_fed_induction_generator).

Now, it's true that below 10% of rated power or so, the efficiency drops, but the amount of additional energy you could gain is really just not worth it. Say you change the turbine from being 66% efficient at 5% of rated power to 90%. You've only increased power output by about 2% of the overall turbine rating, which will have a fairly small impact on annual energy production. I also suspect you would have a hard time with the hydraulic drivetrain concept in getting the full power efficiency up to 90%, so you'll probably lose more energy during high power production than you'll ever gain back in low wind, especially given that you would never want to put a wind turbine on a site where it spends most of its time at <10% power anyways.

Good stuff, thank you. Looking at the graphic below for a 5MW wind-turbine, there is a great deal of cost getting from the turbine to the generator. And, according to this article, the transformer should not be something off-the-shelf but constructed to deal with the variable velocities of the generator.

I propose that using hydraulics as the interim transmission and driving a standard generator at a constant RPM would have similar, if not better efficiency, and reduce much of the upfront cost. 120:1 gearbox alone is between 80-90% efficient.

18-20% of a 5-6 million dollar system is a lot of room to work with.

And there are other things that can be done easily such as regenerative dynamic braking to limit high speeds rather than simply stopping the whole system. That alone could increase overall output in areas that have a lot of extreme wind conditions that require the turbine to be shut down.

It is definitely worth exploring.

 

[Baluncore]

Originally I was looking something small scale like this but then I saw how well this can scale up and spent my time on that. I believe it gets more cost effective at larger scales.

There is not much energy in those typical hydraulic accumulators. Given that 1 unit = 1 kW∙hr = 3.6 MJ
5 gal(us) = 0.018927 m³, 20,000 psi = 137.895 MPa → 2.61 MJ = 0.725 kW∙hour
20 gal(us) = 0.075708 m³, 10,000 psi = 68.947 MPa → 5.22 MJ = 1.450 kW∙hour
50 gal(us) = 0.18927 m³, 3,000 psi = 20.684 MPa → 3.915 MJ = 1.087 kW∙hour

And there are other things that can be done easily such as regenerative dynamic braking to limit high speeds rather than simply stopping the whole system. That alone could increase overall output in areas that have a lot of extreme wind conditions that require the turbine to be shut down.

The hardware will need to be designed to withstand an operational envelope. The maximum energy flow that any part can handle is specified as a power rating in say kilo or megawatts. That specification must include any intended dynamic braking.

During a wind-storm the blades must be feathered, and/or the head rotated side-on, to minimise the rotational speed and total dynamic wind pressure applied to the blades and tower structure. Those peak pressures and speeds will be increased significantly if any attempt is made to oppose them using braking.

A wind-storm may only last for an hour or two but the power rating needed to fight it can be 10 or more times the designed operating envelope. It is better to build 10 units than can duck for cover, than one that will survive a head to head fight with a wind-storm.

 

[deckart]

There is not much energy in those typical hydraulic accumulators. Given that 1 unit = 1 kW∙hr = 3.6 MJ
5 gal(us) = 0.018927 m³, 20,000 psi = 137.895 MPa → 2.61 MJ = 0.725 kW∙hour
20 gal(us) = 0.075708 m³, 10,000 psi = 68.947 MPa → 5.22 MJ = 1.450 kW∙hour
50 gal(us) = 0.18927 m³, 3,000 psi = 20.684 MPa → 3.915 MJ = 1.087 kW∙hourThe hardware will need to be designed to withstand an operational envelope. The maximum energy flow that any part can handle is specified as a power rating in say kilo or megawatts. That specification must include any intended dynamic braking.

During a wind-storm the blades must be feathered, and/or the head rotated side-on, to minimise the rotational speed and total dynamic wind pressure applied to the blades and tower structure. Those peak pressures and speeds will be increased significantly if any attempt is made to oppose them using braking.

A wind-storm may only last for an hour or two but the power rating needed to fight it can be 10 or more times the designed operating envelope. It is better to build 10 units than can duck for cover, than one that will survive a head to head fight with a wind-storm.

Accumulators are really just capacitors. As shown with that calc sheet, they aren't suitable for that scale of storage. Their strength is in that they can absorb and release energy quickly and make hydraulic power available for auxiliary functions.

I think you are right, being that there is a structural capacity of the whole tower assembly and the opposing torque has to be kept safely below that value.

Up to that point, however, turbine speed can be kept at lower speeds than is typical by using hydraulic regenerative braking without any loss of power. A combination of blade pitch and pump control, by bringing individual cylinders on and offline, can keep blade rotation at a low speed throughout much of the higher wind-speed range. Increasing the service life of the main turbine bearing. It may make them quieter too. Not to mention, be safer for wildlife navigating through the blades.

 

[Baluncore]

Up to that point, however, turbine speed can be kept at lower speeds than is typical by using hydraulic regenerative braking without any loss of power. A combination of blade pitch and pump control, by bringing individual cylinders on and offline, can keep blade rotation at a low speed throughout much of the higher wind-speed range. Increasing the service life of the main turbine bearing.

I think you are kidding yourself. The wind energy would need to move along the blades, pass through the pump, twist and lean on the tower, before being removed in the hydraulic oil. The oil would then need to be cooled in a massive radiator. If that was not the case, then it would be operating within the design envelope and generating useful power.

If you are relaxed when you fall, your injuries will be reduced. Likewise, a free-wheeling wind turbine will suffer less damage during a wind-storm than one under any unnecessary load.

It leaves me to question what exactly might you mean by “hydraulic regenerative braking without any loss of power”.

 

[deckart]

I think you are kidding yourself. The wind energy would need to move along the blades, pass through the pump, twist and lean on the tower, before being removed in the hydraulic oil. The oil would then need to be cooled in a massive radiator. If that was not the case, then it would be operating within the design envelope and generating useful power.

If you are relaxed when you fall, your injuries will be reduced. Likewise, a free-wheeling wind turbine will suffer less damage during a wind-storm than one under any unnecessary load.

It leaves me to question what exactly might you mean by “hydraulic regenerative braking without any loss of power”.

Regenerative meaning that the energy is not wasted and expressed as heat. It is put back into the system by increasing the torque of the pumping mechanism, as I have described, to reduce velocity. There are additional methods that can be utilized but that is regenerative braking without any loss of power.

No need to get excited, no one is going to fall, if you don't understand what I'm describing just ask.

 

[cjl]

Good stuff, thank you. Looking at the graphic below for a 5MW wind-turbine, there is a great deal of cost getting from the turbine to the generator. And, according to this article, the transformer should not be something off-the-shelf but constructed to deal with the variable velocities of the generator.

True, the converter is a large chunk of the cost, which is part of the reason for the use (in some cases) of the doubly-fed induction generator I mentioned above. They reduce the cost of the converter considerably, at the expense of a small amount of efficiency and a reduction in flexibility when it comes to reactive power and grid support.

I propose that using hydraulics as the interim transmission and driving a standard generator at a constant RPM would have similar, if not better efficiency, and reduce much of the upfront cost. 120:1 gearbox alone is between 80-90% efficient.

120:1 gearbox alone is between 95 and 97% efficient. There are nowhere near the magnitude of losses you're supposing here. Overall system efficiency is between 85 and 90%, from input shaft power to electrical output.

18-20% of a 5-6 million dollar system is a lot of room to work with.

You're high by about a factor 2 on the cost of a modern 2-3 megawatt machine.

And there are other things that can be done easily such as regenerative dynamic braking to limit high speeds rather than simply stopping the whole system. That alone could increase overall output in areas that have a lot of extreme wind conditions that require the turbine to be shut down.

It is definitely worth exploring.
View attachment 224574

Increasing high wind operational capabilit is worth a negligible quantity of annual energy production, and is severely detrimental to turbine loads. Modern wind turbines already operate out to about 22-27 meters per second wind speed (sustained, not gust), and the time spent above this at most wind generation sites is small enough that it just makes more sense to shut down rather than try to harvest the very small amount of energy at these high speeds. In addition, there's no generator limitation that prevents operation at these high wind speeds - it's purely a loads concern. Speed is already regulated through blade pitch, so excess torque isn't a problem, and most turbines run at a fixed RPM from about 10m/s up to high wind cutout using this blade pitch control method.

 

[cjl]

If you are relaxed when you fall, your injuries will be reduced. Likewise, a free-wheeling wind turbine will suffer less damage during a wind-storm than one under any unnecessary load.

A free wheeling wind turbine will fail in high wind, due to a number of reasons. Higher than nominal blade speed is an emergency, and exceedences of only 30% or so can cause permanent damage. However, modern turbines simply pitch the blades to full feather during high wind, so their rotational speed is very small and there is near zero torque when wind is above about 25m/s or so.