h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

NZ gets alot of its electricity from hydro. One of the big hydro dams (actually two dams on the same river) gets its water from three lakes that have towns on them. One of those towns in particular is prone to flooding.

I've been trying to figure out how much an effect the dam has on the flooding. Here's a map showing the three lakes, their rivers and the dams:

http://maps.google.com/?ie=UTF8&ll=-44.958968,169.302063&spn=1.164186,3.515625&t=h&z=9

Sorry, can't get a better map than that. The big lake in the middle, beside Queenstown, drains east until Cromwell (follow the road). There it meets the river from the top two lakes (Wanaka and Hawea) draining south to Cromwell. At Cromwell one river drains south, through two dams (Clyde is the only one showing on the map) eventually to the sea.

Queenstown is prone to flooding. This happens when there is a big rain when the lake is already high. Where the river leaves Queenstown, it goes through a long narrow gorge. People say this is why Queenstown floods, but I wonder how much the dams have to do with it. The power company artificially manages the flow of all the rivers involved via the dams and control gates at Hawea and Queenstown.

I'm getting varying opinions on this. Some people say that the dams have nothing to do with it because they are so far away, or because the dams let through the same amount of water when full as the river would without the dams (no idea if that is true, but I doubt it). But surely if the point of the dam is to hold back water to use for power generation, then that slows everything down at times.

The way I've been imagining it is a bath full of water. If the drain from the bath is 10cm wide, then water will leave the bath at a certain, determinable rate. But if the pipe beneath the drain gets block in half by a dam, so the gap at that point is only 5cm, won't that slow the rate of water leaving the bath?

Then if at the same time the bath is being filled with the same amount of water that would normally drain through the free pipe, the bath won't drain quickly enough when the 'dam' is in the drainage pipe, and will therefore overflow.

Hope that makes sense.

Well the whole point of a dam is to slow the flow, build up a whole load of potential energy and then use that to generate leccy.

So yes, as you say, of course the flow is slower than if the dam(s) were not there. If it didn't then no lake would form.

Having said that, too much rain will always change the equilibrium level of water and cause a flood. So I wouldn't say the dam was to blame for a flood but is certainly to blame for *where* that flood happens.

'Ang on a mo,

>>the whole point of a dam is to slow the flow, build up a whole load of potential energy...<<
Yep

>>So yes, as you say, of course the flow is slower than if the dam(s) were not there. If it didn't then no lake would form.
<<
Yes, slow it ... initially. Those dams have been there long enough now to have become stable (on average) so the flow must be (on average) the same.
If flooding occurs upstream then the dams can hold back only so much of the excess before they overflow, so presumably, willing or no, they'll have to allow increased flow.


kea, you'll know that the flow starts at Lake Tekapo, at the town of the same name. Have you driven the canal road? it starts nearly opposite where the road down from the Mount John Observatory* crosses Highway 8. It was open to the public until 2008 at least and is a very pleasant drive along a goodly stretch of that make-work canal. It rejoins the main road further South.

which you can visit by arrangement: http://www.newzealandsky.com/earthandsky/ - in case you haven't been there

Yes, that's one of the arguments Rod, that the flow is stable now. Except they change how much goes through the dam all the time. And the point of the dams is to keep the lakes as high as they can get away with. Especially interesting with the floods in Queenstown recently because even though the town isn't flooded, the lake is still at flood height (meaning if they have a big rain again the town will flood). But of course they don't want to let all that water out of the lake because they need it for power generation for the North Island over the winter.

I have driven the canal road, but a very long time ago. Different lakes and dam system


>>
So yes, as you say, of course the flow is slower than if the dam(s) were not there. If it didn't then no lake would form.
<<

The lakes that flood are natural lakes predating the dam. The lake that formed after the dam was built doesn't flood afaik.

The obvious questions would be:

Have you checked how often the town flooded prior to the dam being built as opposed to how many times it's been flooded after?

Has the level of the lake risen since construction of the dam (look at old maps)


I haven't got a detailed map of that town of course, but it does look like Queenstown is partially built on a flood plain at the outlet of a lake... Which, would obviously leave it at risk.

However, if my thought process is correct...

Take this kiddie's experiment...

Your top lake near queenstown is a plastic water bottle.
Your river is a pipe, attached near the bottom of the water container.

Now, the speed the water drains out of the bottle is dependant on the relative heights of the water bottle and the bottom end of the pipe.

Place the bottle on the floor, with the pipe running out from it virtually horizontally, the water will flow out of the pipe relatively slowly.
Put the bottle on top of a wall/door/table/wardrobe and have the bottom end of the pipe at floor level and the bottle will drain much quicker.

Obviously, this is extremely simplified... But as far as I can see, before they built the dam, the river would haved dropped X metres in height before it reached the floor of the valley where they would eventually build the dam.

Now, with the dam in place, the water in the river is actually dropping a lesser distance (X metres - the height of the dam) so it's going to flow slower...

So, theoretically, a downstream dam _could_ have an effect - but, there are so many other variables to take into account, such as the gorge... Is that acting as a choke point? If you repeated the above experiment but put a clamp on the pipe halfway along its length then the flowrate difference between the horizontal pipe (post dam) and a steeper pipe gradient (pre-dam) would be much less.

There are so many other factors that could potentially effect things that it's impossible to say for certain without some very complicated computer modelling.

The power generation company will have complete control over how much water they can release from the lake.

- They can release the same amount as is flowing into the lake, in which case the lake will stay at the same level.
- They can release more than is flowing into the lake, in which case the lake level will drop.
- They can release less than is flowing into the lake, in which case the lake level will rise.

The company will normally plan to have the lake fairly full, but if there is very heavy rain, they will release less water than is flowing into the lake, to reduce flooding below the dam.

If there is extremely heavy rain, the dam will fill, at which stage the power company will be allowing through exactly as water as is flowing into the lake. The effect on flooding below the dam will be exactly the same as if the dam wasn't there.

If there is heavy rain forecast, the company may choose to empty the lake in advice, releasing a lot of water before the storm so that the lake will then be able to hold back floods later. Since the purpose of this is prevent floods, it would be stupid to release so much water in this process that it caused floods.

The only situation in which the power company will contribute to flooding below the dam is if they detect that the dam is structurally unsound and going to collapse. In this case they will release more water that is flowing into the lake, at a time of high flow, to relieve pressure on the dam. I've never heard of any case of this happening.

The flow will not be the same as before.

You're using a lot of that flow energy to turn the turbines and so generate electricity - an I'm sure they will harness as much of it as they can.

So that energy will be taken out of the system and so reduce the flow.


What you describe Rod is what happens when a beaver builds a dam.
When we do it, we take a lot of the energy out of the system.


I would be interested to know what the efficiency of the process is though. It might make little difference to the flow rate...



...but I have a picture of the Grand Canyon in my head and the pitiful remnant of the river that carved it out that now flows at the bottom since they built the dams across the Colorado river.
It seems highly likely to me that very much of the old energy is being harnessed for our purposes.

Hydro efficiency is about 60-65% normally.

In fact that's wrong it's about 80%. Oops!

All this about how the water below the dam can be controlled is all very well, but if I read the original post correctly the flooding in question is a natural lake, -upstream- of a reservoir & dam.

In which case the dam will have nothing to do with it right enough.

that is a bit too fast .. in the original post it said that "The power company artificially manages the flow of all the rivers involved via the dams and control gates at Hawea and Queenstown"

squeezing the control gate at Queenstown could of course make the level in the lake there rise, opening that gate during heavy rains could have too little, too late effects (if the whole thing was dimensioned wrong that is).

On the whole, the lakes act as buffers, over a period of (say) a year all that flows (or rains) in, must flow out. leakage through the soil can usually be neglected, vaporisation can usually be neglected. the buffer lakes will average extremely high or low volumes streaming into the lakes for a short period, and give out a steady stream over a longer period of time.

Control algorithms for the whole thing are rather complex.. including the shape of the lakes, weather forecasts for the next two weeks or so, ... the power company should be able to optimise the flow through the whole system using such control algorithms to set the control gates.

next question would be how to define the optimum .. generator output? generator efficiency? water level at Queenstown?

I'd beg to differ Orcus... Downstream conditions can have an enormous effect on what happens further upstream...

A completely different situation, but similar outcome.

The floods in Carlisle a few years ago weren't simply caused by rain upstream... The levels of rain falling upstream weren't impossibly higher than normal.

The floods were actually caused by a combination of situations... Heavy rain upstream, combined with an extremely high tide downstream and a particular wind and weather combination in the Irish Sea/Atlantic that caused a surge of water to flow into the Irish Sea...

Basically, the quantity of water coming down the rivers hit an obstacle, an extremely strong incoming surge and tide... The obstacle was well downstream of the flooding, but it had the effect of 'backing-up' the water, which ended up flooding a city.

>>
Have you checked how often the town flooded prior to the dam being built as opposed to how many times it's been flooded after?

Has the level of the lake risen since construction of the dam (look at old maps)
<<

Yes, thought of that. I think Queenstown would be prone to flooding even if the dams weren't there. So I'm not suggesting that the dams are the single cause of flooding. I don't know historically if there are more floods after the first dam or the second dam than before either, but weather patterns have changed enough to make that not particularly useful.

I don't think the lake level was raised. The dam is a long way down stream. I'm not even sure how much the control gates affect the lake, but I suspect not alot.

I did wonder about the rate of drop and how that would influence things too, but wasn't sure how relevant that was over such a distance.

Orcus : "...but I have a picture of the Grand Canyon in my head and the pitiful remnant of the river that carved it out that now flows at the bottom since they built the dams across the Colorado river."

I'd always been given the impression that that was largely down to the extent of Agricultural take, rather than effect of damming the river.

Unless you actually remove the water from the river, the slowdown effect from creating a dam will only last until you have filled your lake. Then, if you don't want to overtop the dam, you'd need to increase the flow, to match the input + losses through evaporation etc.

>>
If there is heavy rain forecast, the company may choose to empty the lake in advice, releasing a lot of water before the storm so that the lake will then be able to hold back floods later. Since the purpose of this is prevent floods, it would be stupid to release so much water in this process that it caused floods.
<< Gnomon

There are flooding issues downstream from the dam (in rivers), but I was talking about the lake town which is upstream by many miles.

The weather events that cause flooding aren't particularly well predicted. If they were then presumably the power company could empty the lake enough to prevent flooding. You would need alot of notice for that though. The lake at Queenstown takes a long time to drop to normal levels once it's flooded - weeks not days.

The level of the lake does not affect what happens upstream, except to a very small extent by influencing the level of the water held underground (the water table). It really depends on how high the town is above the lake (in altitude).

>>The level of the lake does not affect what happens upstream<<

I'm not sure why you've said that. The lake is at the top of the catchment, the town beside the lake. The dam is many miles downstream.

The town is at the same level as the lake.

In that case, flooding in the town _is_ dependent on the level of water in the lake, even though the town is above the lake (in altitude).

When the lake level is high, the level of water underground (the water table) will be correspondingly high. When rain falls on the town, it will not be absorbed into the ground, but will sit on the surface, or run off as a flood.

When the lake level is low, the level of water underground will also be low, and rain falling on the town will soak straight into the ground, not sitting on the surface or running off.

The flooding is completely dependent on the lake level, not sure why you would think otherwise but maybe I wasn't clear. Rain falling in the town is not an issue. It's the rain in the lake catchment that fills the lake and floods the town.

The part of the town that floods is the same level as the lake, possibly even lower in places (there's a wall at the bottom of the main street I think between the road and the town. When it floods they have to sandbag that end of town to stop the lake from inundating those streets and shops.

Here's a vid of the last flood a few weeks ago, showing the wall and sandbagging.

http://www.3news.co.nz/Queenstown-locals-label-flood-phoney/tabid/309/articleID/153584/Default.aspx