An update on what is happening at Guri dam

March 11, 2010

Well, I have good news and bad news. The good news, is that the model that I presented the other day is doing ok so far, it was predicting for yesterday a height of the water level at Guri dam of 253.25 meters and the reported level in the Opsis website is 253.16 meters. This is not bad, given that I used data up to February 15th., thus, after 14 days, the error is only 9 centimeters, which tracks the model quite well. Cross your fingers.

Except the bad news is that this is more complicated that it may have seemed at first sight.

First of all, it has been raining! That is definitely good news. Here is a plot of the water flow into the Guri dam:

as you can see, the flow almost doubled after February 25th. as the rain brought in more water for a few days, going from around 400 cubic meters per sec. to 800. Then the flow dropped again to around 600 m^3/sec. and on March 8th. it rained again and water inflows levels are around 700 m^3/sec.

The bad news is, that despite the fact that water inflows are higher than in all of February, the water level is still dropping, as seen below in the the plot of the Height above sea level of the water:

As can be seen, the water level keeps dropping. We have had more rain and more water, but the rate at which the level has been dropping seems to be continuing.

How come?

Easy, as the water into the dam has increased, the flow out through the turbines that produces the electricity has also been increased as can be seen in the next chart:

As you can see, as the rain has provided more water, the dam has been used more and the flow out has increased. In fact, it seems to be increasing daily as if the people running it are controlling it accordingly.

This implies, that despite the rain the daily drop in height has not slowed down, in fact, it has increased:

almost reaching 20 centimeters daily.

Thus, I am not sure why it is dropping faster, it would seem as if more water is being allowed to flow to produce more electricity, but the rate of increase in the flow out does not seem to be as large as to justify the faster drop in the height. It may be that because it is much hotter and there is a less surface area to the lake, there is much more evaporation or the shape is playing a role.

Comments are welcome, particularly by the experts, who may enlighten us on dam management!


51 Responses to “An update on what is happening at Guri dam”

  1. This is really attention-grabbing, You are a very professional blogger. I have joined your feed and look ahead to looking for more of your wonderful post. Additionally, I’ve shared your site in my social networks

  2. cat Says:

    im fat :(((((((
    how do i lose weight

    im trying fat free nip but i eat to much of it then cough up a hair ball

    any advice??????

  3. cat Says:

    hello im bobby catnip

  4. moses Says:


    Last night I tried a 3rd order polynomial using Excel 2000 and got the following:

    Predicted level
    Date Square Cubic

    25-Mar-10 250,95 250,76
    26-Mar-10 250,80 250,60
    27-Mar-10 250,64 250,44
    28-Mar-10 250,49 250,28
    29-Mar-10 250,33 250,11
    30-Mar-10 250,17 249,95
    31-Mar-10 250,01 249,78
    1-Abr-10 249,85 249,62
    2-Abr-10 249,69 249,45
    3-Abr-10 249,53 249,28
    4-Abr-10 249,37 249,11
    5-Abr-10 249,21 248,94
    6-Abr-10 249,05 248,77
    7-Abr-10 248,88 248,60
    8-Abr-10 248,72 248,43
    9-Abr-10 248,56 248,25
    10-Abr-10 248,39 248,08
    11-Abr-10 248,22 247,90
    12-Abr-10 248,06 247,72
    13-Abr-10 247,89 247,55
    14-Abr-10 247,72 247,37
    15-Abr-10 247,55 247,19
    16-Abr-10 247,38 247,01
    17-Abr-10 247,21 246,82
    18-Abr-10 247,04 246,64
    19-Abr-10 246,87 246,46
    20-Abr-10 246,70 246,27
    21-Abr-10 246,53 246,09
    22-Abr-10 246,35 245,90
    23-Abr-10 246,18 245,71
    24-Abr-10 246,00 245,52
    25-Abr-10 245,83 245,33
    26-Abr-10 245,65 245,14
    27-Abr-10 245,48 244,95
    28-Abr-10 245,30 244,76
    29-Abr-10 245,12 244,56
    30-Abr-10 244,94 244,37
    1-May-10 244,76 244,17
    2-May-10 244,58 243,98
    3-May-10 244,40 243,78
    4-May-10 244,22 243,58
    5-May-10 244,04 243,38
    6-May-10 243,85 243,18
    7-May-10 243,67 242,98
    8-May-10 243,48 242,77
    9-May-10 243,30 242,57
    10-May-10 243,11 242,36
    11-May-10 242,93 242,16
    12-May-10 242,74 241,95
    13-May-10 242,55 241,74
    14-May-10 242,36 241,53
    15-May-10 242,17 241,32
    16-May-10 241,98 241,11
    17-May-10 241,79 240,90
    18-May-10 241,60 240,68
    19-May-10 241,41 240,47
    20-May-10 241,22 240,25
    21-May-10 241,02 240,04
    22-May-10 240,83 239,82

    The difference is not much at the beginning, then it starts to spread out. Understand that in 2002-2003 they went down to 244 meters. We will be there on the first of May, if the rains do not start soon.

  5. Mick Says:

    Since shallow lakes are warmer than deep lakes and rainfall is going to be the same temperature as the current surface temperature(70-90 degrees Farenheight), I would think that the average temperature of the Guri is increasing. This would increase evaporation rate wouldn’t it?

  6. tinerfe Says:


    If you use a 3rd order polynomial (in Excel: add trend line -> polynominal -> increase default order from 2 to 3) then you get nearly the same results as the power-law model with exponent 1/7 until about 2010-05-25. The 2nd order polynomial deviates clearly from the power law model at about 2010-04-01.

    The reality already rejects the 2nd order polynominal: if you plot the residual errors between the 2nd order polynominal and the real data you see a clear curvature (the real data tend to fall faster than the model).

    There is up to today no notable difference between the 3rd order polynominal model and the power-law model with exponent 1/7 and the real data, but physical thinking favors the power-law. And the power-law has only 3 free parameters (T = day of total run-down; a = exponent; const = proportionality factor) whereas the 3rd order polynominal has 4 free parameters (“Occam’s razor” favors simpler models with less free parameters).

    My current best parameters are (t measured in days, h in meters above 127 m a.m.s.l.):

    T = ‘2010-07-09 00:00:00’ + 0.821 days
    a = 1/7.0
    const = 63.61

    h = const * (T-t)^a

    This gives this prediction (mean error +- 0.022 m since Jan 1st):

    date, at 23:59:59 h
    height of Guri surface above mean sea level, in meters
    daily drop, in meters

    16.03.2010 252,26 0,16
    17.03.2010 252,11 0,16
    18.03.2010 251,95 0,16
    19.03.2010 251,79 0,16
    20.03.2010 251,63 0,16
    21.03.2010 251,47 0,16
    22.03.2010 251,31 0,16
    23.03.2010 251,14 0,16
    24.03.2010 250,98 0,17
    25.03.2010 250,81 0,17
    26.03.2010 250,64 0,17
    27.03.2010 250,48 0,17
    28.03.2010 250,30 0,17
    29.03.2010 250,13 0,17
    30.03.2010 249,96 0,17
    31.03.2010 249,78 0,17
    01.04.2010 249,61 0,18
    02.04.2010 249,43 0,18
    03.04.2010 249,25 0,18
    04.04.2010 249,07 0,18
    05.04.2010 248,89 0,18
    06.04.2010 248,70 0,18
    07.04.2010 248,52 0,19
    08.04.2010 248,33 0,19
    09.04.2010 248,14 0,19
    10.04.2010 247,95 0,19
    11.04.2010 247,75 0,19
    12.04.2010 247,56 0,20
    13.04.2010 247,36 0,20
    14.04.2010 247,16 0,20
    15.04.2010 246,96 0,20
    16.04.2010 246,76 0,20
    17.04.2010 246,55 0,21
    18.04.2010 246,35 0,21
    19.04.2010 246,14 0,21
    20.04.2010 245,92 0,21
    21.04.2010 245,71 0,21
    22.04.2010 245,49 0,22
    23.04.2010 245,28 0,22
    24.04.2010 245,05 0,22
    25.04.2010 244,83 0,22
    26.04.2010 244,60 0,23
    27.04.2010 244,38 0,23
    28.04.2010 244,14 0,23
    29.04.2010 243,91 0,23
    30.04.2010 243,67 0,24
    01.05.2010 243,43 0,24
    02.05.2010 243,19 0,24
    03.05.2010 242,94 0,25
    04.05.2010 242,69 0,25
    05.05.2010 242,44 0,25
    06.05.2010 242,18 0,26
    07.05.2010 241,92 0,26
    08.05.2010 241,66 0,26
    09.05.2010 241,39 0,27
    10.05.2010 241,12 0,27
    11.05.2010 240,85 0,27
    12.05.2010 240,57 0,28
    13.05.2010 240,29 0,28
    14.05.2010 240,00 0,29

  7. moses Says:


    That is an interesting model which takes into account the physics of the reservoir.

    I did a trend analysis which uses a 2nd order polynomial using data from 2-Dic-2009 until two days ago.

    Here are my estimates, using the polynomial and the solver add in of Excel:

    250 meters: 31-Mar
    248 meters: 12-April
    245 meters: 29-April
    240 meters: 26-May

  8. Curacao Says:

    A second attempt at the table of the earlier post

    ————Average level ————- net average ———— cm/d level drop
    ————drop p/d ——————- daily outflow ———– for every 1000
    ————[cm] ————————-[m3/min] —————–m3/min net outflow

    Jan ————10.8 —————————3596 ————————– 3.00
    Feb ————12.8 —————————3960 ————————- 3.23
    March(1-14)–14.7 —————————4218 —————————3.48

  9. Curacao Says:


    If you perform some calculations on the Opsis data there is actually pretty clear evidence that your impression that the gradual increase in daily average level drop is not only caused by an increase of net out flow of water but also by an increase in level drop at the same outflow.

    Average level net average cm/d level drop
    drop p/d daily outflow for every 1000
    [cm] [m3/min] m3/min net outflow

    Jan 10.8 3596 3.00
    Feb 12.8 3960 3.23
    March(1-14) 14.7 4218 3.48

    The last column in effect means that in March the level is dropping on average 16% faster than in January, for the same net out flow.

    Another interesting fact from he Opsis data is that the level drops about 3 cm more on a Monday than on the Sunday before. This 3 cm in a way can be considered an estimate for the effect that an extra day of closing of all businesses would have (a Monday would become a Sunday). Over a 3 month period that would save about 30-40 cm. Does not sound like a lot, at 15 cm/d level drop that would give you less than 3 days of ‘time gain’. If on the other hand it starts raining and inflow comes closer to outflow this could quickly become more than a week, potentially just enough at the end of May or beginning of June.

  10. CVC Says:


    I’ve been collecting and tracking the reservoir and national electric data and publishing it periodically.

    here you can see the daily inflows, releases, power generation, lake levels etc.

    The factors most significant are:
    rainfall…historical variability is very wide, and current inflows are spectacularly low. A reversion to the mean is inevitable but unpredictable. El niño normally has about a 15% suppressive effecton rainfall.
    Thermo…..very little progress has been made in raising output thusfar.
    Heat…hot weather has significantly raised power demand in March as it typically does.
    Rationing…this is already reducing demand a lot. Strict enforcement and further curtailment could slow the decline more, but demand has been spiking back up since march 1 as it got hot.

  11. tinerfe Says:

    I suggest a better model which so far (since January 1st, 2010) predicts perfectly the water level of Guri (the residual error is just noise). It takes into account that the storage level ‘h’ (Guri level above sea minus the level of the lower waters at the exit of the turbines at 172 m a.m.s.l.) should depend on a simple power law if there is a continuous energy extraction from the lake and a constant inflow):

    h = const x (T-t) ^a

    where ‘(T-t)’ denotes the remaining time until the (purely theoretical) total run-down of the lake to the lower water level below the dam and ‘a’ an exponent. It happens that the exponent a=1/7 does fit the reality perfectly (this is to my surprise, I thought that it was 1/4 for a conical sea bed, but it seems that the shores get steeper when the sea surface decreases and the efficiency of the turbines drop faster than simply due to the loss of pressure).

    So here is the empiric formula for the storage level above the lower waters (in meters above lover waters):

    h = 63.6 x (‘2010-07-09’ – t)^(1/7)

    The level of Guri in meters above mean sea level is hence:

    H = h + 127 = 63.6 x (‘2010-07-09’ – t)^(1/7)

    where ‘t’ stands for the current day.

    Some examples:

    1. Water level on t=’2010-03-14′ (yesterday, Sunday, 24:00h):
    There were 117 days remaining until 2010-07-09, take the
    7th root from 117 and multiply this by 63.6. Add 127.
    Result: 252.58 m. (Error: 0.01 m).

    2. Water level on t=’2010-05-15′ (day of predicted collapse):
    There will be 56 days out until 2010-07-09, take the 7th root from
    56. Multiply the result by 63.6.
    Result: 240.03 m

    This model tracks the reality with a mean error of 0.020 m and a maximum error of 0.04 m since the first day of this year.

    The predicted collapse will be, according to this model, on May 15th (240 m a.m.s.l.). 244 m will be reached on April 29th. 248m will be reached on April 10th.

  12. moses Says:


    Check this Corpoelec presentation:

    On slides 11,12 and 13 you can see that if Guri gets to 240 meters, you have to take out 5000 Mw of generating offline. It also says very clearly on no. 13 that 240 meters is the minimum level for operation.

  13. Mérida Says:

    As a side not 11th only lost 15cms, so while over all the trend is up, it seems to have averaged around 2-3cms per month depending on rain fall. 13cms in February around this time, and 10cms around this these time in January.

    I’m going to assume that this may be to the two factors presented in your artical.

    I’m guess most days next month if rains continue being around the same level as we have now we will see 17cms, if less rain then around 18cms. We won’t see 20cms till around may 11th of May if rain continues to remain the same.

  14. Mérida Says:

    Chavez recently said that the level the Guri dam can go to is 236m. Anyone have any idea if this is true or false?

    This may explain their lack of urgency on the matter. Maybe that can buy them another 2 months, not like Chavez said, another year, which I think is just BS.

    How does one check on this? Who first stated 240 was the limit?

  15. Moses Says:


    We reached similar conclusions using different methods. What surprises me is the regression factor, almost a perfect fit.

    When it starts to rain heavily there will be an inflection point, bad for this regresion but good for all of us !

  16. moctavio Says:

    You did a fit to the height h as a function of time using a polynomial:


    In my earlier post (with fewer points) I did a fit to the dailt heigth change, but linear


    so, it is the same in the end

  17. moses Says:

    No data today, I think they post one day behind . . .

    I am no expert on data analysis, but I used a tip found here:

  18. moctavio Says:

    That’s similar to what did, because I did a linear fit to the derivative, no?

    BTW, no data today..

  19. moses Says:


    Here is the link to the page:

    Look for RESUMEN DIARIO DEL SEN it takes some time to load. There is a dialog box for the date. If you select more than 5 -10 different dates you get a //20 message and you have to close the page and try back later.

    Direct link:

    Good Luck !

  20. Mérida Says:

    Thanks for all the information you guys have provided. Where do you see the level of the water?

    I went to the Opsis link up top, but I had a hard time finding the data, if anyone can link me exactly where I can see the level. My spanish is not as hot as I would like it to be, so any help is appriciated.

  21. moses Says:


    Taking data from Opsis, from 2-Dic-09 to 11-Mar-10 of Date vs. Height, and using a 2nd degree polynomial, I get the following:

    Correlation Coefficient: 0.99989 (pretty close to 1)

    Date Level

    29-Mar-10 250.35
    12-Apr-10 248.08
    26-Apr-10 245.69
    24-May-10 240.49

    Biggest difference I get between predicted and actual data is 0.07 mts.

    If it does not start to rain soon, probably by mid April the water flow through the turbines will be throttled down and rationing will hit Caracas . . .

  22. J. Suarez Says:

    Ladies, Gentlemen.

    The reasons for the acute drop are quite simple, the numbers do not add up:

    1. More generation, not less from the Guri Dam and Bajo Caroni keeps being demanded, which in turn requires more water flow in a dimisnishing conical shape reservoir, aggravated by heat on a smaller lake and increased evaporation losses (e.g. water that can not be turbinated to generate power/MW.)

    2. Why more generation and not less is being demanded from such a National Treasure and Referential Engineering World Power houses. Several reasons:

    2.1 It is a matter of physics:
    Height of the water column x Gravity force x Water flow (inflow of the river) x the efficiency of the individual turbines being used tell us the power that can be produced.
    The less water column you have the more water you need to make the same amount of MW.
    2.2 Then comes the reality of why more power is being demanded. This is no brainer, it is simple: POLITICS for the here and now and forget the long term consequences.

    The simple matter is that the thermoelectric fleet that has to show up and be counted on at this critical time, is in great state of disrepair.

    Most remaining units do not have the required availability and reliability to be counted on, then the Guri and Lower Caroni plants (Guri, Caruachi, Macagua I, II and III) need to generate more power or there would be more power rationing on a very short calendar year, which is politically incorrect.

    2.3 This is a technical problem but the POLITICS for the here and now and forget the long term consequences, requires clarification: The year 2010 for Venezuela does not end on December 31, 2010, but rather on September 26, 2010; does not let capable engineers deal with the facts.

    2.4 The numbers do not add up 516 “New MW” since Jan 1, 2010, list them and show them to the nation. At best only 145 MW new MW can be accounted for or then tell the nation and prove beyond any reasonable doubt that the OPSIS daily reports are wrong.

    2.5 Do not confuse activity with accomplishments, do not lie even though a plant “may be inaugurated” it is probably far from being properly released for reliable commercial operation. Be more honest with the Nation.

    2.6 Several units are late and the state of the pre/operating trails are showing unexpected problems like Planta Centro Unit 1.

    2.7 Show more respect and appreciation for the Venezuelan workers in the electrical plants and the engineers that know their units, value them, listen to them.

    I am not a politician, and my comment is not politically motivated, the crisis is neither of the right or the left, it affects a whole nation that deserves much better than current results.

    Venezuela, needs big chunks of power fast from the thermoelectric plants in short order, 11 or 30 MW at a time helps but the pace is not fast enough.

    Planta Centro needs to be recuperated as soon as possible and be repowered but even though practical proposals have been given no one in a position of accountability has the will to do anything about it.
    The ironic thing the parts are in the country but “other priorities and agendas” of personal reason or gain win over the legitimate right of the people to have safe, reliable and efficient power or electricity.

    3.0 Before I forget, when will we get to 240 MASL water line?

    That is a complex problem, that takes many variables difficult to control, which therefore requires a mathematical model to predict hard to control variables:

    3.1 The weather (when will it start raining and how much),
    3.2 The availabity/reliability of the Lower Caroni plants, especially that of the older Units in Guri’s Powerhouse 1: Units 1 to 10, their time is coming to carry the the nation like they did between 1968 to 1980.
    But remember, this grand old lady is only capable of less than 3.000 MW and only about 2.200 cubic meters per second, which would severely limit the MW production downstream plants.
    3.3 The availabity/reliability of the thermoelectric plant fleet,
    3.4 The up to date volumetric capacity curves of each of the segments of the entire reservoir segments e.g. useful volume:
    3.4.1 – Zone A (Top and largest water tank) From 271 MASL to 240 MASL.
    3.4.2 – Zone B (Middle and second largest water tank) From 240 to 215 MASL. A tank that is more of a factor for Units 1 to 10 than for Units 11 to 20. At 235 MASL we can only use 2 units from Powerhouse II.
    3.4.3 – Zone C (Lower and smallest water tank) From 215 MASL to about 200 MASL, the water intake of Units 1 to 10 is at 196.5 MASL.

    At that point the reservoir is much smaller less than 800 Square Km and you can not use this water to make power, it is the dead volume of the reservoir.

    3.5 How much power rationing the politicians can stomach in light of the great hardship put on the people of the interior of Venezuela, since Caracas is of the options. “Caracas is Caracas lo demás es monte” rings true again. This is not about “regionalism” this is about reality, “what you do speaks so loud I can not hear what you say”.

    Because of the complexity, it is hard to predict, at best the answer has to be given in a probabilistic terms, but one thing is for sure, if we keep the Strenuous generating pace of this plants:

    It is not on April 6, 2010 as reported once and it is not on the 3rd week of June 2010.

    The Conical shape of the reservoir would not allow for that, it is more like Mid may or third week of May 2010, again simple physics:

    At 271 MASL we have a lake that is 3.911 Square KM, the lake is getting smaller by the day and there is less water in it. Be honest with the people you have a lot of meters left but these meters do not equate to the same volume of useful “turbinable water”.

    Also do not stretch reality, be honest:

    Caruachi and Macagua I, II and III need to see 6.000 cubic meter per second to reach their max. power output, if Guri can not give that volume we can not make the power.

    Also the above plants have reservoirs, but these reservoirs can not be used as the Guri’s Dam reservoir, these have smaller useful volume and the lower height of their dams require a lot of flow to make the power due to the conversion factor of their installed units.

    I am not a politician, and my comment is not politically motivated, the crisis is neither of the right or the left, it affects a whole nation that deserves much better than current results.

    I write as a concerned engineering professional which only politics is to produce results for the benefit of mankind.

    You can decree a National Electrical Emergency, you can not decree: Integrity, Technical Competence, Transparency, Planning

    May be start with Moral y Luces son Nuestras Primeras Necesidades and know the difference between Responsibility and Accountability.

    I love you tons my Great Venezuela.

  23. Joaquin Rodriguez Says:

    En primer lugar queria decir que, por los datos que nos presentas, es evidente que el embalse Guri esta perdiendo mas agua que la que le esta entrando por sus fuentes naturales. El que no incremente su nivel a pesar de que se ha aumentado un poco el aporte de natural en estos ultimos tiempos se puede explicar, si consideramos que el descenso del nivel del embalse provocado por la sequia anterior produjo una disminución de la energia potencial de la columna de agua. Para compensar esa perdida de energia y seguir produciendo igual cantidad de corriente eléctrica se necesita aumentar la cantidad de agua que se hace pasar por las turbinas . Ese exceso de agua no está siendo compensado por el aporte de las fuentes narturales y el nivel del embalse continua descendiendo…La cosa empeorará mas rápidamente en la medida que la altura de la columna de agua ,en el área de máquinas, siga disminuyendo lo que puede llevar al colapso de la producción de energia eléctrica, si no se producen grandes lluvias en la cuenca que la abastece en los próximos tres meses . Las lluvias no muy pronunciadas lo único que pueden provocar es que la planta siga funcionando a baja capacidad por un tiempo con la esperanza de que se produzcan lluvias copiosas que restablezcan el embalse. El restablecimiento del embalse no va a ser fácil pues se le está sacando agua en grandes cantidades continuamente pues la planta no se puede parar dado que ella sóla produce alrededor del 70 % de toda la energia electrica de de Venezuela.

  24. Moses Says:

    Sorry for the repeated posts…. Hope you enjoyed the article !

  25. Roberto N Says:

    That was a good article moses, thank you.

    As a kid I camped near the Caroni. An Edelca employee explained to us that to dam the river when construction began, 18-20 ton rocks were used, and were swept away.

    They then resorted to filling railcars that weighed 45-50 tons, and tossing those in.

    Those did the trick.

  26. Mike Kramer Says:

    There is one more aspect regarding dams, which has not been mentioned yet in the discussions. Every river causes erosion and the generated sediments are deposited in hydro-electric basins due to the reduction of flow velocity. My friends tell me that the river is black, an indication of a high solids content. Over time these sediments fill up the basin and can limit the lifetime of the project (see attached link).

    The recent rapid increase in water drop level appears to indicate that a significant section of the lake has a current basin depth around the 255 m level (level drop = net offtake / active surface area). When this falls dry the active surface area is reduced and even with a slight decrease of net offtake due to starting rains, the level drop increases.

    Prudent management of 70% of the country’s electricity generation capacity would require periodic monitoring of the basin depths to forecast the lifetime of power generation at Guri. If anyone has access to this historic and current basin depth data, both this year’s level drop and the dams lifetime can be easily calculated.


  27. Charly Says:

    Moctavio, “to avoid the cavitation problem”. Who says that they want to avoid it. In fact, in an electoral year, it make sense to carry on generating and damn the consequences. This government is hoping from one crisis to another one, so by the time the rains are back another crisis will need attending. This is what I love about this country and why I am staying here. Every day bring new challenges and new fun. The perfect place for a libertarian, try that in the good old US of A.

  28. Mérida Says:

    I have a couple of questions that if anyone knows anything about please reply.

    1. What happens when we reach 240 meters? Does all power from Guri go out? Do we simply loss a percent? Please explain.

    2. The other dams furthur down the stream, how will they be effected, if lets say Guri stops letting water go? Do they have a good deal of water, and are they working full, or can they pick up some of the slack for Guri?

    3. These thermogenerators he is putting all over Mérida, not sure where else, a chavista told me no pollution is pumped out of them. This seemed a complete crock of excrement, as while some filiters will block out sulfer oxides, you still have CO and CO2 that is hard to filter out. Anyone know anything about these generators. I’ve not done any research and am being lazy.

    4. Both Columbia and Brazil are offering power, any news on how much they are offering, and if it would be enough for all of Venezuela? Also can the goverment of Venezuela afford it?

  29. Charly Says:

    AG, this is a vortex, probably created by the running of turbines. It does not look too bad. However, when the eye of the vortex increases, air is captured by the vortex and goes down to the turbines and hit the blades at high velocity under high pressure. The metal of the blades get pitted and the turbine efficiency deteriorates, a phenomenon known as cavitation. This is what happens when turbines run at low head such as would happen at Guri under critical reservoir level. One has two alternatives, stop turbining to avoid turbine damage, then there is no power production. The other alternative is to keep on turbining and producing power, let the units deteriorate and once the critical period is over, rehabilitate the turbines, an expensive proposition especially considering that at Guri, several units are already down on maintenance. This is a call for Chavez, but the poor man is such a pussy anyway.

  30. moses Says:

    Check this link (first part) of the history of Guri Dam:

  31. AG Says:

    Can someone explain what this picture means?

  32. torres Says:

    It seems that a diminishing head (i.e., water pressure) would not lead to a *linear* relation to compensation water flow. As head goes down, the amount of compensation should have to increase, I imagine, exponentially.

    As to rain, it also seems that the longer time passes without refilling, the more the ground would absorb whenever it rains, and the more would be lost to capillary action of the ground due to evaporation from the ground.

  33. moses Says:

    Check this nice article about Guri Dam (First part)

  34. moses Says:

    What it is sad about this situation is that there seems to be orders to keep the Guri dam generating electricity, no matter what the consequences in the near and middle term to the dam may be. (Damn the dam ?)

    If the rationing measures had been taken in September – October last year we wold have been in a much better shape. Today the may have slowd down a bit the decreasing water level.

    Want to read a little history behind the Guri Dam ? See the first article in this link of Water Power Magazine:

    Default Story Type
    Roll of honour
    09 July 1999

    When construction draws to a halt, and even after a dam has been inaugurated, we are perhaps all guilty of not fully appreciating the individual achievements of each hydro scheme. Suzanne Moxon decided to give members of the dam construction industry the opportunity to talk about their favourite project — one which they think should be considered a feat of construction. Other opinions are invited.

    (Guri Dam)

    Located on the Caroni river in Venezuela is one of the largest operating hydroelectric projects in the world. The 10,300MW Guri hydro power scheme, according to Vito DiCosola, senior partner with Harza Engineering Company, should be included in the dam construction industry’s roll of honour. Built over a 23-year period, Guri was completed in 1986 after utilising a unique construction concept — development of the project was phased to minimise initial investment and to match power supply to the growth in electrical system demand.

    ‘The phased development allowed us to build this project economically,’ DiCosola explained. ‘Other concepts floated at the beginning were to build the dam to the final height at the first stage. This obviously would have been a much easier design but the region would not have absorbed all of the power, making it difficult to finance the scheme.

    ‘Harza proposed building a smaller project and then, when power demand increased, we planned to raise the dam accordingly. This was a unique concept when suggested over 40 years ago,’ says DiCosola. ‘It was also very difficult to do, and at the time had never been carried out on such a large scale before. We raised the dam by 52m.’ Planning for this project proved to be a very complex job. DiCosola explained that the design of Guri dam had to acknowledge the interface between the old and new concrete when increasing its height. ‘We needed to know exactly how to do this at the very start of the job. In 1963 the first construction phase started and by this time we had to have decided how to interface the future dam extensions. There was a lot of planning involved but it also meant that the final raising of the structure caused fewer problems.’ Innovative ideas continued to appear throughout the construction of Guri dam. A significant concept was developed which replaced post-cooled concrete. ‘Usually with post-cooled concrete,’ DiCosola explained, ‘a pipe is embedded in the concrete layers and circulates cooling water, which reduces the temperature of the concrete. As we were obviously very concerned about thermal cracking when raising the dam, we decided to use ice chips instead of the cooling water. This proved to be very successful and eliminated the need for cooling pipes.’ Other potential problems for Harza included river diversion and control during construction. Seasonal flow range varied from a low of 5000m3/sec to a wet season high of 12000m3/sec, and there was not a constant PMF. ‘One of our concerns was the control of the river,’ DiCosola says. ‘We had to give contractors early warning if a major flood was imminent so EDELCA, the project owner, invested money in weather forecasts and set up an automatic rainfall and gauging station in the upper drainage basin to monitor the situation. Fortunately, a major flood did not occur and the spillway performed very well.’ Designing the Guri spillway was an intricate operation. It was divided into three separate chutes to permit sequential raising of the crests with two chutes operational while one was under construction. Three floating bulkheads were used to close one spillway chute at a time when raising the spillway crests. At 20x36m, these were the largest bulkheads used for this purpose in the world. Indeed, according to DiCosola, during construction, many world records were established at Guri. These included:

    •Highest raising of an entire dam in one step (by 52m).

    •Largest amount of concrete placed in one month (200,000m3).

    •Most fill placed in a 24-hour period (234,500m3).

    •Largest construction trestle with eight 800ton whirley cranes for placing concrete with 6.1m3 buckets.

    •Largest dead-end transmission tower supporting two 800kV transmission lines.

    •Largest number of monitoring instru-ments installed at one site (1800).

    As Ienan Prusza, chief civil design engineer at EDELCA, explains some of these records have now been broken but he still considers Guri to be a marvel of dam construction in the twentieth century. ‘The construction of the third and final stage was an accomplishment,’ he said. ‘It included raising the spillway and gravity dam by 52m, constructing the second powerhouse and excavating the second tailrace channel to increase the generation capacity from 2000-10,300MW.’ Notable construction methods worthy of mention, Prusza believes, include the pre-wetting of the borrow area by sprinklers which allowed fill placements of more than 200,000m3/day. The successful introduction of aeration in the spillway also acted as protection against cavitation and erosion of the concrete due to sustained high flow conditions (average annual flow of 5000m3/sec and PMF inflow of 48,100m3/sec).

    ‘Furthermore,’ Prusza went on to add, ‘the compact and incremental design of Guri resulted in a very economical power plant with power costing US$550 per kW.’ As DiCosola pointed out, the power was so inexpensive at the time that it attracted aluminium and iron ore smelting industries into the area.

    Construction management and co-operation between all parties involved is a very important aspect of any project. DiCosola was keen to point out that there was excellent co-operation between Harza and EDELCA and that they worked very well together on many different aspects of the project. This resulted in the project being ‘pretty much’ on schedule and budget, and being hailed as a success by Harza and EDELCA.

    Maintaining a focus on the Caroni river in Venezuela, Ienan Prusza also considers the Macagua project to be a construction achievement. Located about 10km upstream from the Caroni river’s confluence with the Orinoco river, the 2540MW project was completed in 1996.

    ‘At this site the river widens to about 4km and divides into many channels, cascading down to the level of the Orinoco,’ Prusza said. ‘Some of these channels terminate in the spectacular falls of Cachamay and La Llovizna, one of Venezuela’s most scenic attractions. Also, because of the project’s location within the Ciudad Guayana urban zone (an industrial city with a growing population of 700,000 inhabitants) the project represented a construction challenge from an environmental point of view. The cost of the mitigation measures implemented to address human and environmental impacts of the project represented 14% of the total cost of the scheme.’ The environmental construction characteristics of Macagua are summarised below:

    •A low head powerhouse was constructed for the sole purpose of providing a permanent flow of 660m3/sec to the Cachamay and La Llovizna waterfalls. Additionally, this plant generates 160MW of power.

    •Artificial lakes, covering an area of 1.4km2, were constructed to stabilise the water level in the surrounding parks.

    •An eight-lane highway was incorporated into the project to connect the cities of Ciudad Guayana, Puerto Ordaz and San Felix. It provides access for visitors to the surrounding park areas.

    •The reservoir has controlled beach areas for use by the population.

  35. concerned Says:

    I must be confused. I thought the excuse for the rationing was to allow a reduction of generation, thereby conserving water level. Not increasing the flow to compensate for head pressure to maintain the pathetic energy supply generated by the few turbines still mechanically able to operate.

    In a normal situation with fully fuctioning turbines before the dry season, a rationing program would actually equate to a reduction of power generation and a conservation of water above the dam.

    You guys are spending too much time trying to analyze this as if it were a normal situation and the only figures needed were water heights and flows…in other words, real data that is trendable and predictable. You will never get an accurate prediction due to the human factor involved and the politics of the situation. As crazy as it sounds, Chavez needs this to dam to go below critical levels and interrupt power services. He is not interested in your welfare…just covering his ass and distracting away from other current events. Most importantly, hiding the REAL truth about the failures of the electrical sector after being taken over by the government.

  36. moses Says:


    You can visualize it as generated power =fwater flow x head (height). To maintain the same generated power, if you decrease the height, you have to compensate with additional flow. But this has a limit since at high flows you risk running into cavitation, which causes erosion and damage to the turbines.

  37. JoE Says:

    Great articles! But… they’re getting a bit too long. A few were TL:DR.

  38. Charly Says:

    Power generation (Megawatts) is directly proportional to turbine flow (meter cube per second) and head on the turbines (meters). Head is the difference between current reservoir level and level of the river at turbine restitution less more or less constant friction losses.

    For simplification, the reservoir can be viewed as a cone, hence the top centimeter of reservoir contains more water than the bottom centimeter. Also, the head at the top of the reservoir is higher than at the bottom.

    For a given power production, one needs to drawdown the reservoir further at low levels to obtain the same flow as there is less volume per unit drawdown (eg. centimeter) than at higher reservoir levels. So this is the first reason why the level drop keeps increasing as the reservoir goes down. Each successive centimeter of water contains less water than the preceding centimeter so the shortfall is made up by using more centimeters to obtain the required turbine flow.

    In addition, as the reservoir goes down, the head also goes down and should be compensated by additional flow hence additional drawdown on the reservoir.

    If it rains, volume is added to the reservoir and the level goes up. For a given precipitation amount generally measured in centimeters, volume added by rainfall decreases as the reservoir goes down because of the shape of the reservoir.

    For a given evaporation amount also measured in centimeters, evaporated volume decreases as the reservoir goes down because of the shape of the reservoir.

  39. moctavio Says:

    Dagoberto: I have that number, but it is a % of what? That is one number I dont quite understand (Thus, I dont talk about it :-))

  40. dagoberto Says:

    Miguel, thanks for the good work!

    Let me suggest you to focus your model NOT on water level, BUT on usable volume.

    Usable volume is a figure given by the OPSIS as a percentage. If you could estimate the absolute value, and couple it with you inflow/outflow data, I believe you’ll be able to get a better model that is independent of Guri lake basin shape, leading to increased accuracy.

    Indeed, if you get an independent absolute value of usable volume, you could even model the evaporation rate.

    I hope this helps.

  41. IvoSan Says:

    as “torres” said, flow has to be increased to compensate for the decresing pressure.

  42. HalfEmpty Says:

    I’m surprised this data isn’t treated as a state secret.

  43. Federico Rivero S. Says:

    Two thoughts or aspects that you may want to take into consideration in your analysis:
    1. In line with Samaniego’s comment and regardless of its contour the lake is a basin. As it shrinks the same volume takes up more depth. Ergo, the water level drops faster.
    2. As the depth drops the incline of the lake’s shores also drops, tends to become more leveled. And with les depth the combined effects of heat absorption in the water surface and the lake’s bottom add up to account for faster evaporation.
    The basin effect, can be easily determined If we know the drop rate of the lake’s shores. However, even thou I am convinced that the evaporation speed increases as the depth decreases; I have no clue as to how to measure it.
    Any way Thanks for your analysis; it does not calm my anxieties, but keeps the old brain occupied in more rewarding matters.

  44. torres Says:

    not an expert, but I venture a guess that as water pressure goes down with the water level, flow must be increased to compensate.

  45. moctavio Says:

    Joel: I dont either unless they are trying to keep it that way, can they?

  46. moctavio Says:

    Moses: But why then allow more flow, I dont quite get that, no?

  47. joel Says:

    Does It seem like the plot from the 15th to the 9th is really linear? I don’t think they could have done that if they tried.

  48. moses Says:


    If the critical level is 248 – 245 meters, we will get there in the first week of April (248) or last week of April (245).


    Check your mail for details.


  49. Enrique Samaniego Says:

    Hola Miguel, the reason is the same as a paper water cone, with a hole at the bottom. As the water comes out the paper cone, water level drops down at a faster rate. Do the experimented yourself and you’ll see. Same theory is with the dam. Highest surface area low rate drop, smaller surface area highest rate drop.

  50. Roger Says:

    Im no expert just a research assistant! But several days ago I when to this site which uses google maps and images and I could see the contours and water when the picture was taken. I just went there and what I see is a big black area and the published map from years ago. I checked Caracas and even Miraflores has no black spot over it. thus I have a concern about the data.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: