Power shares

[Snoopy]

Unlike the other Gentailers, Contact's hydro is not really much of a battery at all. Clyde's storage is only 6 GWh, and at normal inflows that can be emptied in 22 hours. Roxburgh is even less at 1.6 GWh and can be emptied in as little as 6 hours. The only true storage on the Clutha scheme is at Hawea, but that has such severe ramp rate and flow restrictions that changes in outflow must be made 15 - 24 hours before the water is either required, or no longer required, at Clyde. In addition, the minimum flow requirements downstream of Roxburgh mean that the Clutha stations can only be used to buffer a maximum of 200 MW of intermittent generation such as wind.

1.6GWh (storage capacity at Roxburgh) = 1,600MWh

The Roxburgh dam can provide 1,600MWh/200MW = 8 hours of equivalent wind energy storage (using Jantar's figures).

Roxburgh has a maximum generation capacity of 320MW. So the 1,600MWh storage capacity will be emptied in:

1,600MWh / 320MW = 5 hours (maximum discharge rate). Of course, hydro does not run flat out all the time. A 60% utilisation rate would see that the storage at Roxburgh empties in:

5 / 0,6 = 8.333 hours,

which very neatly matches your 200MW Wind Energy buffer. Jantar.

But if the output of the Clyde dam feeds the Roxburgh dam, and Clyde is of larger capacity, why can't Clyde replenish Roxburgh as it empties (should that be required)?

For Clyde, 6,000MWh / 200MW =30 hours

So really you can store 30+8 =38 hours of wind energy in the Clyde/Roxburgh system.

Jantar you say:
"The only true storage on the Clutha scheme is at Hawea, but that has such severe ramp rate and flow restrictions that changes in outflow must be made 15 - 24 hours before the water is either required, or no longer required, at Clyde."

But if the storage in Roxburgh/Clyde is kept at at even half of its maximum level (19 hours of storage) and water is short, that means there is enough time to bring the storage at Hawea into play should the coupled wind energy projects not deliver. So why is it not realistic to regard Hawea, Clyde and Roxburgh as part of a giant Clutha River battery?

The Lake Onslow scheme, if built to capacity will be 1200 MW. The good part about that is that unlike conventional hydro, pumped storage can buffer twice its own capacity in wind generation. So that 1200 MW at Onslow would permit a further 2400 MW of installed wind generation. That is sufficient to replace our current thermals, plus allow for some overbuild for future years load increase.

Isn't that theoretical buffer capacity for windfarms at Onslow modelled around the wind generation capacity being on or off? As more wind farms spread up around the country, isn't it less likely that they will all be generating together, or, on the other side of the coin, be becalmed together?

SNOOPY

[Carpenterjoe]

The Lake Onslow scheme, if built to capacity will be 1200 MW. The good part apout that is that unlike conventional hydro, pumped storage can buffer twice its own capacity in wind generation. So that 1200 MW at Onslow would permit a further 2400 MW of installed wind generation. That is sufficient to replace our current thermals, plus allow for some overbuild for future years load increase.

Curious janter,

Have you ever thought out side the box and looked at the mines around nz as potential water batteries? Genex in Australia a close to doing it in queensland.

Macras has some big holes and the geographical nature of the location has pleanty of potential. 500m above sea level 100m+ pits.

Waihi has a few issues, land stability, location ect.

Few old coal mines on the west coast with plenty of elevation.

I know they would be alot smaller. But could be good use of old infrastructure and provide grid stability and back up.

[Jantar]

1.6GWh (storage capacity at Roxburgh) = 1,600MWh

The Roxburgh dam can provide 1,600MWh/200MW = 8 hours of equivalent wind energy storage (using Jantar's figures).

Roxburgh has a maximum generation capacity of 320MW. So the 1,600MWh storage capacity will be emptied in:

1,600MWh / 320MW = 5 hours (maximum discharge rate). Of course, hydro does not run flat out all the time. A 60% utilisation rate would see that the storage at Roxburgh empties in:

5 / 0,6 = 8.333 hours,

which very neatly matches your 200MW Wind Energy buffer. Jantar.

But if the output of the Clyde dam feeds the Roxburgh dam, and Clyde is of larger capacity, why can't Clyde replenish Roxburgh as it empties (should that be required)?

For Clyde, 6,000MWh / 200MW =30 hours

So really you can store 30+8 =38 hours of wind energy in the Clyde/Roxburgh system.

Jantar you say:
"The only true storage on the Clutha scheme is at Hawea, but that has such severe ramp rate and flow restrictions that changes in outflow must be made 15 - 24 hours before the water is either required, or no longer required, at Clyde."

But if the storage in Roxburgh/Clyde is kept at at even half of its maximum level (19 hours of storage) and water is short, that means there is enough time to bring the storage at Hawea into play should the coupled wind energy projects not deliver. So why is it not realistic to regard Hawea, Clyde and Roxburgh as part of a giant Clutha River battery?



Isn't that theoretical buffer capacity for windfarms at Onslow modelled around the wind generation capacity being on or off? As more wind farms spread up around the country, isn't it less likely that they will all be generating together, or, on the other side of the coin, be becalmed together?

SNOOPY

Sorry Snoopy, it isn't that simple. The 6 GWh storage figure for Clyde is allowing for the water to be used at both stations.

Roxburgh has a minimum discharge of 250 cumecs, which means for practical purposes it has to generate on 3 out of its 8 generators at 120 MW just to meet that resource consent.

Clyde has a minimum flow of 120 cumecs in daylight hours, but can shut from 1 hr after sunset to 1 hour before sunrise measured at a flow station at the Clyde golf course. Because of flow times, and the need to keep Roxburgh topped up it effectively means that Clyde can shut down for a maximum of 5 hours and usually less. When running each machine is generating between 100 and 108 MW, depending on the head and tailwater levels, but can reduce to 70 MW or go up to 116 MW, but that means running inefficiently and effectively spilling water. The 200 MW buffering available from the Clutha is the mean difference between its base offered load at both stations and the efficient running of additional plant. If the dams are kept at mid range as you suggest that is 11 hours of storage over and above normal running.

Now here is where it starts to get difficult. The Clutha stations are Run of River, which means their base offers to the market is the amount of water flowing into Clyde. There is not a normal flow as the Clutha is the most volatile hydro catchment in the country.


As for Onslow being modelled for wind farms being either on or off, that is not the way it has been modelled either. Rather it is assuming that the wind has a load factor of 35 -40%, and the model applied a random walk around that number.

[Jantar]

Curious janter,

Have you ever thought out side the box and looked at the mines around nz as potential water batteries? Genex in Australia a close to doing it in queensland.

Macras has some big holes and the geographical nature of the location has pleanty of potential. 500m above sea level 100m+ pits.

Waihi has a few issues, land stability, location ect.

Few old coal mines on the west coast with plenty of elevation.

I know they would be alot smaller. But could be good use of old infrastructure and provide grid stability and back up.

Pumped storage for dry year cover requires a high head, large storage, steep ground, and a water supply at the lower reservoir. If it is just for buffering then the storage only needs to be measured in days, not months. Mines, if at a high enough altitude, could be use as an upper reservoir, but the volume of water they can hold is usually quite small.

Macraes is 500 m above sea level, but would get water and discharge back to the Shag river upstream of Dunback. As the full depth is not available for storage, (only around half of it) the size of the pit would only hold 14 MWh of energy in storage. That would be an expensive project for not much return. Grid scale Tesla batteries would be a better option than small pumped storage schemes.

[Snoopy]

Now here is where it starts to get difficult. The Clutha stations are Run of River, which means their base offers to the market is the amount of water flowing into Clyde. There is normal flow as the Clutha is the most volatile hydro catchment in the country.

Jantar I can't make sense of the above bit. I presume the above contains a typo and is meant to say:

"There isn't a normal flow as the Clutha is the most volatile hydro catchment in the country. " ?

SNOOPY

[Jantar]

Jantar I can't make sense of the above bit. I presume the above contains a typo and is meant to say:

"There ism't a normal flow as the Clutha is the most volatile hydro catchment in the country. " ?

SNOOPY

Correct. My typo. Fixed in the original post now.

[Snoopy]

As for Onslow being modelled for wind farms being either on or off, that is not the way it has been modelled either. Rather it is assuming that the wind has a load factor of 35 -40%, and the model applied a random walk around that number.

I am trying to understand better the concept of mixing and matching wind generation and hydro generation to make maximum use of the natural air flow and water flow to produce a stable combined power generation system.

There are certainly seasonal droughts for hydro inflows. We are too much of a farming nation for low water events not to hit the headlines. But periods of low wind do not hit the headlines in the same way. I had assumed that wind generation availability was dependent on what weather system was moving over the country at the time, and at a seasonal level that was largely predictable. But weather systems move over the country in just a few days. So I had assumed there was no seasonal variation of wind over a longer period and that wind generators simply worked around individual weather system peaks and troughs. What I have just related may not be correct, and I am prepared to have my views revised.

Rain is obviously a weather event too.. But water falling in the highlands may take days or even months (if the water falls as snow) to permeate the river systems. So I have always considered hydro as the 'stable' part of a wind/hydro generation partnership. To this end I am somewhat flabbergasted to learn that wind energy input is modelled as

"assuming that the wind has a load factor of 35 -40%, and the model applied a random walk around that number."

From an annual all enveloping wind energy viewpoint that sounds about right. But in the context of a hydro system with a total hydro storage capacity of just a few hours....

Clyde's storage is only 6 GWh, and at normal inflows that can be emptied in 22 hours.

.....then assuming a load factor of 30-40% at any particular time looks like it will be almost certainly wrong and by a wide margin, My admittedly anecdotal experience is that when the wind blows it does so for several hours, while likewise calm periods last for several hours. So surely if you are planning to integrate wind energy with hydro on a daily basis, and your total hydro-storage capacity is measured in hours , then you have to plan around extreme events (virtually no wind or close to 100% utilised wind supply)?

SNOOPY

[Jantar]

…. My admittedly anecdotal experience is that when the wind blows it does so for several hours, while likewise calm periods last for several hours. So surely if you are planning to integrate wind energy with hydro on a daily basis, and your total hydro-storage capacity is measured in hours , then you have to plan around extreme events (virtually no wind or close to 100% utilised wind supply)?

SNOOPY

Replace that "hours" with "hours to days". The duration of wind depends on the type of weather patterns prevailing at the time.

An anticyclone will mean very light, or even no wind, but with a possibility of sea breezes for wind farms close to the coast. Sea breezes are rather unpredictable, as they can be forecast as to whether or not they will happen, but not the direction, actual strength, or distance of fetch inland. Anticyclones typically take 3 days or more to pass over.

A cold front will be preceded by building Nor-westerlies for 1 to 3 days before it hits, but as it passes over the wind will drop and back to the Sou-west before picking up again. Then wind speed will fall away as the pressure builds. It is this change as the front passes that is when wind changes are measured in hours rather than days.

A warm front will be preceded by very light winds from almost any direction except NW, and veer towards a more northerly influence after it passes. This process also takes 2 or 3 days

Then there are occluded fronts, stationary fronts, tropical storms, polar vortices etc all with differing characteristics regarding wind strength and duration.

Add all these together and it means forecasting wind generation on very long time scales is quite easy, but shorter durations are very prone to error. It is common to see the wind forecast to the market to be out by well over 100 MW, and even 200 MW discrepancies are not unusual. The worst I have seen was a 390 MW error, and our installed capacity for wind at that time was 679 MW.

[Carpenterjoe]

Pumped storage for dry year cover requires a high head, large storage, steep ground, and a water supply at the lower reservoir. If it is just for buffering then the storage only needs to be measured in days, not months. Mines, if at a high enough altitude, could be use as an upper reservoir, but the volume of water they can hold is usually quite

Thanks Jantar,

Your definitely thinking big picture.
Personally not a fan of battery storage, their lifespan is far to short. I think small pump storage projects are vital for grid risk management and insurance.

Have you looked at Genex and their 250mw ((2000mwh (max 8 hours)) pump storage project?

https://www.genexpower.com.au/project-details.html

Should cost about 650m.

[Jantar]

...

Have you looked at Genex and their 250mw ((2000mwh (max 8 hours)) pump storage project?

https://www.genexpower.com.au/project-details.html

Should cost about 650m.

Yes, and we already have a suitable site and power station in NZ that could give us very similar results at a much cheaper price.

Tokaanu power station was originally designed to able to be used as pumped storage between Rotoaira and Taupo. However it never received consent to use a suitable operating range at Rotoaira which was limited to 30 cm, instead of the 1 m it needed.

The station exists, both lakes are already there. All it needs is for the machines to be retrofitted and get resource consent (good luck with that one) and IWI approval (not a chance in Hades). Total plant cost around $120 M.