The ultimate infrastructure project: A Transcontinental aqueduct to save the American Southwest from becoming a desert.

The American Southwest has always been subject to drought cycles, some worse than the one that is now devastating the area. Below is a very interesting presentation from ASU about a previous civilization in the Phoenix area, thriving and then gone.

Arizona State University presentation

Will it happen again?

The problem:

  1. Lake Mead will be emptied in less than 10 years with the current usage pattern. Then what?
  2. The hydroelectric power from Lake Mead (and Lake Powell) is diminishing as the lakes are emptied.
  3. the aquifers in Arizona, especially in the Phoenix and Tucson area, and to some extent New Mexico and the dry part of Texas are being drawn down and are at risk of being exhausted.
  4. The Salton Sea in the Imperial Valley of California is maybe the most polluted lake in all of U.S.A. It is even dangerous to breathe the air around it sometimes. The area contains maybe the largest Lithium deposit in the world.
  5. The Colorado River water is too salty for good irrigation .
  6. The Colorado river no longer reaches the Gulf of California. Fishing and shrimp harvesting around the Colorado River Delta is no more.
  7. 40 million people depend on the Colorado River for drinking water. The population is still rising rapidly in the West. Will they have water in the future?
  8. Except for California there is not much pumped Hydro-power storage in the American Southwest.
  9. Texas has plenty of wind power, but no pumped hydro-power storage. This makes it difficult to provide peak power when the sun doesn’t shine and the wind doesn’t blow. Nuclear power is of no help, it provides base power only. Peak power has to come from coal and natural gas plants.
  10. New Mexico has some ideal spots for solar panels, but no water is available for pumped storage.
  11. Arizona has a surging population, wind and solar power locations are abundant, but no pumped hydro-power storage.

The solution:

Build a transcontinental aqueduct from the Mississippi River to the Colorado River capable of transporting 15 million acre-ft of water yearly through Louisiana, Texas, New Mexico and Arizona. It will be built similar to the Central Arizona Project aqueduct, supplying water from the Colorado river to the Phoenix and Tucson area, but this aqueduct will be carrying seven times more water over five times the distance and raise the water more than twice as high before returning to near sea level. The original Central Arizona Project cost $4.7 billion in 1980’s money, the Transcontinental Aqueduct will cost around $340 Billion in 2021 money applying simple scaling up principles.

The Mississippi River has a bad reputation for having polluted water, but since the clean water act the water quality has improved drastically. Fecal coli-form bacteria is down by a factor of more than 100, the water is now used all the way down to New Orleans for drinking water after treatment. The lead levels are down by a factor of 1000 or more since 1979. Plastic pollution and pharmaceutical pollution is still a problem, as is the case with most rivers. The Ph is back to around 8 and salt content is negligible. Mississippi water is good for irrigation, and usable for drinking water after treatment.

But the aqueduct will do more than provide sweet Mississippi water to the thirsty South-west, it will make possible to provide peak power to Texas, New Mexico and Arizona. In fact, it is so big it will nearly triple the pumped Hydro-power storage for the nation, from 23 GW for 5 hours a day to up to 66 GW.

The extra pumped hydro-power storage will come from a number of dams built as part of the aqueduct or very adjacent to it. The water will be pumped from surplus wind and solar power generators when available. This will provide up to 20 GW of power for 5 hours a day. If not enough extra power has been generated during the 19 pumping hours, sometimes power will be purchased from the regular grid. The other source of pumped hydro-power storage is virtual. There will be more than 230 MW LFTR (Liquid Fluoride salt Thorium Rector) power stations strategically stationed along the waterway providing pumping of water for 19 hours and providing virtual hydro-power output.

These 43 GW of hydro-power capacity will be as follows: Louisiana, 0.4 GW; Texas, 18,5 GW (right now, Texas has no hydro-power storage, but plenty of wind power); New Mexico, 10.5 GW; Arizona 13.6 GW. In Addition, when the Transcontinental Aqueduct is fully built out, the Hoover dam can provide a true 2.2 GW hydro-power storage by pumping water back from Lake Mojave; a 3 billion dollar existing proposal is waiting to be realized once Lake Mead is saved.

The amount of installed hydroelectric power storage is:

U.S. operating hydroelectric pumped storage capacity

Most hydroelectric pumped storage was installed in the 70’s. Now natural gas plants provide most of the peak power. This aqueduct will double, triple the U.S. pumped peak storage if virtual peak storage is included. By being pumped from surplus wind and solar energy as well as nuclear energy it is true “Green power”. Some people like that.

What follows is a description of each leg of the aqueduct. Each leg except legs 9 and 10 ends in a dam, which holds enough water to make each leg free to operate to best use of available electricity and provide peak power on demand.

Leg 1: Atchafalaya river (Mississippi river bypass) to Aquilla lake, a distance of 360 miles.

Leg 2: Aquilla lake to Brad reservoir (to be built), a distance of 100 miles.

Leg 3: Brad reservoir to North of Baird dams. (to be constructed), a distance of 60 miles

Leg 4: North of Baird dams (to be constructed) to East of Sweetwater dam (to be built), a distance of 60 miles.

Leg 5: East of Sweetwater dam (to be constructed) to Grassland Canyon Lake (to be made), a distance of 50 miles.

Leg 6: Grassland Canyon Lake (to be made) to White Oaks Canyon Lake (to be made), a distance of 110 miles.

Leg 7: White Oaks Canyon Lake (to be made) to the Arch Lewis Canyon Lake via a 20 mile tunnel under the Guadaloupe Mountains in New Mexico.

Leg 8: Arch Lewis Canyon Lake to Martin Tank Lake, a distance of 50 miles.

Leg 9: Martin Tank Lake to Poppy Canyon Reservoir, a distance of 210 miles.

Leg 10: The Poppy Canyon Upper and Lower Reservoir. A Hydro-power storage peak power plant.

Leg 10, alternate solution: Poppy Canyon Reservoir to Cove Tank Reservoir, a distance of 13 miles.

Leg 11: Poppy Canyon Reservoir to San Carlos Lake, a distance of 80 miles.

Leg 12: San Carlos Lake to the Colorado river following the Gila river, a distance of 280 miles.

The Transcontinental Aqueduct. Leg 12: San Carlos Lake to the Colorado river following the Gila river, a distance of 280 miles.

Stage 12 is a true delivery of water on demand aqueduct. The San Carlos lake has a storage capacity of a million acre-ft, the ideal buffer from the peak power demand driven uphill stages to the major delivery stage. San Carlos lake is now mostly empty, but will be normally filled to 85% of capacity, slightly less in advance of the winter snow melt. The Lake would look like this:

San Carlos lake, about half full

The Coolidge dam is now decommissioned, the lake is too often empty and the dam suffered damage in the power plant and it was no longer economical to produce power. The retrofitted dam will have a power generation capacity of up to 19,000 cfs the top of the dam is at 2535 ft, the typical water level is at 2500 ft and the drop is 215 feet, giving a maximum power output of 315 MW.

Coolidge dam before rebuilding

From there the stream follows the Gila River all the way to the Colorado River with the following drop-offs:

Where the Arizona central project waterway crosses the Gila river it will deliver up to 500 cfs to Tucson

Where the Gila river meets the Salt river it can deliver up to 1,500 cfs to the Phoenix-Scottsdale metropolitan area.

To the Martinez lake it can deliver up to 15,155 cfs, the design capacity of the All American canal. This will of course be nearly always far less, dependent on the need for water for irrigation, but we dimension the aqueduct to accommodate maximum flow. The Martinez lake is puny, and would easily be overwhelmed by surges in the water flow. To accommodate this, the Senator Wash Reservoir will have to be upgraded to be able to pump up or down at least twice as much water as is it’s present capacity. Lake Martinez is at about 180 feet elevation, and Senator Wash Reservoir is at a maximum elevation of 240 feet.

The Martinez lake and the Senator Wash Reservoir.

The rest of the Transcontinental Aqueduct empties out where the Gila river joins the remainder of the Colorado river a few miles downstream. It will be able to carry up to 6, 000 cfs of water to accommodate the needs of Mexico and also provide a modest amount of water to assure the Colorado river again reaches the ocean, maybe restoring some shrimp fishing in the ocean.

The 1944 water treaty with Mexico provides Mexico with 1.5 million acre-ft per year, more or less dependent of drought or surplus. It will be increased only on condition that when the Transcontinental aqueduct is finished, the New River in Mexicali will be cut off at the border, and Mexico will have to do their own complete waste water treatment.

There will be water allocated to the Salton Sea. Proposed will be the world’s largest Lithium mine, mining the deep brine, rich in Lithium. (about a third of the world supply according to one estimate). This requires water, and as a minimum to allow mining in the Salton Sea the water needs to be cleaned. This requires further investigation, but the area around the Salton Sea is maybe the most unhealthy in the United States.

The maximum power generating drop during this last leg will be (2500 – 190 – 2.2X 280) = 1694 feet. With an average flow of 14,000 cfs this will generate 1.8 GW of power, but the realized power output will be determined by the actual water demands.

Climate change: IPCC report is ‘code red for humanity’. Not so fast. The Arctic ice sheet and Greenland ice is doing quite well, thank you. A Limerick.

The ice in the Arctic will stay

In Greenland it snowed every day

New white snow, what a sight!

Reflects back all the light

No climate change here, this i say.

Back in 2012 the Greenland ice sheet had an unprecedented melt, and the prediction was that all the Arctic ice would be melted in September of 2015, having reached the Climate tipping point from which there is no return to a normal climate unless we reorganized society into a more totalitarian global governance.

Well, the tipping point didn’t happen, so hopefully global governance will not happen either, even though many are trying.

These are the latest charts for arctic temperatures, ice and snow for August 15:

The ice-pack on Greenland has been melting much less than normal during the melting season

And yesterday’s snowfall over Greenland

Remember, H2O is a condensing gas, when cooled off it condenses into clouds.

Clouds cool by day and warm by night, a one percent difference in cloud cover means more than the increase in CO2.

The only place this doesn’t work is in deserts, if no clouds form Forget CO2, but let us not make any more deserts. The American South-west is in danger of being “desertified” unless we restore the fragile water balance in the region.

The Transcontinental Aqueduct. Leg 11: Poppy Canyon to to San Carlos Lake, a distance of 80 miles.

Stage 10 was a true pumped hydro-storage peak power stage, producing up to 11.0 GW electric power for up to 5 hours a day. In stage 11 the flow will be a maximum flow of up to 25,000 cfs, but with periods of less flow during low electricity demand, all to accommodate both water needs and power demands.

The power generating drop is on average (4320 – 2510 – 80×2.2) = 1,634 feet. This stage is capable of generating maximum 2.4 GW of power during all times.

San_Carlos_Lake is located within the 3,000-square-mile (7,800 km2) San Carlos Apache Indian Reservation, and is thus subject to tribal regulations. It has been full only three times, in 1993 it overflowed the spillway and about 35,000 cfs of water caused erosion damage to natural gas pipelines. The lake contains now (April 6,2021) less than 100 acre-ft of water. All fish is dead.

When former President Coolidge dedicated the dam in 1930, the dam had not begun to fill. Humorist Will Rogers looked at the grass in the lake bed, and said, “If this were my dam, I’d mow it.”[

When the Transcontinental aqueduct is built the lake will always be nearly filled, level will be at 2510 feet with flood control nearly automatic, it will never overflow, and it will look like this:

The San Carlos lake, when filled will hold 1,000,000 acre-ft of water.

The Coolidge dam will have to be retrofitted for a 25,000 cfs water flow

The Transcontinental Aqueduct. Leg 10: The Poppy Canyon Upper and Lower Reservoir. A Hydro-power storage peak power plant.

Stage 9 ended up near the Poppy Canyon Lower dam. The aqueduct will release water to the dams when necessary to compensate for evaporation and seepage losses, but it will otherwise be independent of the aqueduct. The upper dam is 400 feet high and will top out at 5360 feet with a maximum water level at 5350 feet. The lower dam is 480 feet high, and the water tops out at 4670 feet. This stage can deliver peak energy only 5 hours a day, or deliver the day’s worth of peak energy whenever called for. To make this possible there will be a tunnel and pumping station capable of delivering up to 95,000 cfs when called for. The drop is maximum (5360 – 4200) = 1160 feet and minimum (4950– 4670) = 280 feet with an average of 720 feet, delivering 5.3 GW of pumped hydro-storage peak power for 5 hours a day for a total of 26.5 GWh. For the other 19 hours a day water will be pumped up from the lower dam to the upper dam, requiring sixteen 100 MW LFTR power stations, or 30 GWh. the difference is because of the 93 percent efficiency in the turbines and generators. But for 5 hours a day the 16 LFTR’s will produce 1.6 GW of virtual peak power.

Here as always, the preferable power to lift the water will be produced by excess solar and wind power. But when the sun doesn’t shine and the wind doesn’t blow. the power has to be available. When excess power is available these LFTR plants are free to produce hydrogen, to be stored and used for more peak power

Lake Oroville going dry, but why?

Once full was the Oroville Lake.

Now empty. How much does it take?

No pumped power galore

for the windmills to store.

Blame Climate change, not your mistake.

In 2016 then Governor Jerry Brown declared that California was in a permanent state of drought, so they might as well atart to prepare for water rationing.

The Lake Oroville Dam had a large crack in its spillway, and it was part of the regular maintenance to fix it, but since they were in a permanent drought the lake would never again be full, so there was no need, and certainly no hurry to fix it. Then in 1917 it started to rain again, the lake started to overflow, and instead of a less than 20 million maintenance task it became an over 1 Billion dollar rescue effort with helicopters trying to dump stones in the eroded parts of the dam

That was in early 2017. then in early 2019 it was full again, and with proper conservation measures there was enough water for 5 to 7 years with normal rainfalls from then on, so not to worry. Look where the lake levels are now:

Lake Oroville water level 8 8 639.67 feet

power. But that is not allSince today’s level is below the intake for hydro-power there will be no power from Oroville dam until spring melting season, thus depraving California about 440 MW of power. But that is not all, it also eliminates Oroville Dam of 117 MW power as a Hydro-power storage “battery” for excess wind power, so more wind turbines will have to be shut off when the wind is blowing since there are no customers for excess power. On the other hand, when the wind is not blowing it will have to be replaced by coal or natural gas, which are in insufficient supply. The future is full of brownouts, and rotating blackouts.

This is how the Oroville Lake looks now:

What is most galling is that of the water released in March of this year, before farmers really started to use water, 91% of the released water went into the San Francisco Bay to save the Delta Smelt, a totally useless fish, but protected. For the moment I can not think of a more inept way to run a water and energy business.

The Transcontinental Aqueduct; Will it pay for itself?

The goal of the Transcontinental Aqueduct is to save Lake Mead, save the American Southwest from becoming a desert, provide Hydroelectric peak storage for Texas, New Mexico and Arizona, provide sweet Mississippi water for irrigation, provide water to the Colorado river so it again can reach the ocean, revitalize San Carlos lake, provide more and better drinking water to 30 million people, to name just a few benefits.

The cost is substantial. The biggest problem is that the aqueduct must be substantially completed at full capacity before any benefits from the water will materialize. The cost to bring the aqueduct to half capacity is 300.5 billion dollars in construction cost only. This includes the cost of half the pumps or generators needed for full capacity, but not the cost of the power plants. Add to this the cost of filling the aqueduct and the 11 dams. The aqueduct itself will contain 1 million acre-ft of water when filled, the 11 dams will contain about 800,000 acre-ft when half full. To pump 1.8 MAF an average of 5000 feet requires about 10 TWh, when losses are included. Th filling stage water will be pumped, using excess wind and solar power at bargain rates, about 4 c/kwh , the same as the LFTR will produce when fully installed. This is about 320 million dollars in “liquid investment” The electric cost of moving one acre-ft from the Mississippi to the Colorado River is 6 MWh. This power is initially bought from off-peak wind and solar power, but as the aqueduct is completed with true hydropower storage up more and more the power will be generated with 100 MW LFTR power plants, the hydropower storage will be filled with excess wind and solar power.

In short: assuming a 50 year amortization plan for the aqueduct, and money available at 2%, , it will cost 12.5 billion a year in capital cost to deliver 7.5 MAF water from the Mississippi River to the Colorado river or any point in between, or $1,670 per acre-ft. Add to that $240 for electricity and another $50 per acre-ft in overhead and maintenance, the cost will be $1960 per acre-ft

When the aqueduct is fully built up, it will cost $13.4 billion yearly in capital cost to deliver 14.5 MAF of water from the Mississippi River to the Colorado river or any point in between, or $ 925 per acre-ft. The other costs stay the same, so the total cost of water will be $ 1,215 per acre-ft.

I have not yet mentioned the other major benefit of the Transcontinental Aqueduct. If I wanted the lowest cost of water possible, I would have used the lower route, going through the Texas lowlands to El Paso before routing it through New Mexico and Arizona. I routed it through the high and dry parts of Texas and New Mexico, at extra altitude penalty. The intent is to provide Hydropower storage at select places. These places are ideal for wind and solar power, but they need to store the energy when the sun is not up or doesn’t shine, or the wind doesn’t blow. Right now that is provided by coal and natural gas. Conventional nuclear power is best for use as base power only, so this transcontinental aqueduct will provide up to 23 GW of pure hydropower storage for 5 hours a day, but the LFTR nuclear stations providing the energy pumping the water in the aqueduct will shut off the pumps for five hours a day, or when the need arises, and instead provide another 20 GW of virtual hydropower power.

These 43 GW of hydropower capacity will be as follows: Louisiana, 0.4 GW; Texas, 18,5 GW (right now, Texas has no hydropower storage, but plenty of wind power); New Mexico, 10.5 GW; Arizona 13.6 GW. In Addition, when the Transcontinental Aqueduct is fully built up, the Hoover dam can provide a true 2.2 GW hydrostorage poser by pumping water back from Lake Mojave, a 3 billion dollar existing proposal waiting to be realized once Lake Mead is saved.

The amount of installed hydroelectric power storage is:

U.S. operating hydroelectric pumped storage capacity

Most hydroelectric pumped storage was installed in the 70’s. Now natural gas plants provide most of the peak power. This aqueduct will double, triple the U.S. pumped peak storage if virtual peak storage is included. By being pumped from surplus wind and solar energy as well as nuclear energy it is true “Green power”. Some people like that.

The Transcontinental Aqueduct. Leg 9: Martin Tank Lake to Round Mountain Rockhound Reservoir, a distance of 210 miles.

Stage 8 went from Arch Lewis Canyon Lake to Martin Tank Lake. The Martin Tank Lake dam is 22260 feet wide and 230 feet high. The Lake will contain about 30,000 Acre-ft when full, about eighteen hours worth of storage.

The elevation at the Martin Tank lake will top out at 4620 feet with maximum water level at 4610 feet. The aqueduct will first descend to 3980 feet, as it crosses the Rio Grande in La Mesa, a distance of 50 miles. The elevation difference is (4620 – 3980 – 50 X 2.2) feet = 530 feet. Releasing 24,000 cfs of water 530 feet will generate 1 GW of energy for 19 hrs/day. From La Mesa it will climb to the Round Mountain Rockhound Reservoir. The dam is 400 feet high and will top out at 5360 feet with a maximum water level at 5350 feet. The total lift of the water in stage 9 is (5350 – 3980 + 160×2.2) feet = 1722 ft. To lift 24,000 cubic feet per second 1722 feet requires seven 500 MW LFTR nuclear reactors, . T he Round Mountain Rockhound Reservoir will contain about 40,000 Acre-ft of water when full, about one day of storage. For 5 hours per day these seven reactors used in this stage can provide 3.5 GW of peak power to the grid instead of pumping water, thus acting as a virtual hydroelectric peak power storage. The Round Mountain Rockhound Reservoir will look like this:

The Transcontinental Aqueduct. Leg 8: Arch Lewis Canyon Lake to Martin Tank Lake, a distance of 50 miles.

Stage 8 ended in Arch Lewis Lake. It will be filled mostly during the 5 hours of peak power generation. During the other 19 hours the fill rate will be very low leading to lowering water levels.

It has a 3000 feet wide and up to 480 feet high dam, topping out at 4620 feet, and the lake holds a volume of up to 60,000 acre-ft of water.

From the Arch Lewis Lake dam to the Martin Tank Lake the distance is 60 miles the way the aqueduct takes. It will first descend to 3720 feet before rising to 5190 feet. The descending drop is (4620 – 3720 – 2.2 x 9), or up to 890 feet. The Martin Canyon Lake will top out at 5200 feet with maximum water level at 5190 feet. The total lift of the water in this stage is (5190 – 3720 + 51×2.2) feet = 1582 ft. To lift 25,000 cubic feet per second (1582 x 1.03 – 890 x 0.97) = 766 feet requires three 500 MW LFTR nuclear reactors plus the use of 80 MW from the grid. The Martin Tank Lake dam is 22260 feet wide and 230 feet high. It will contain about 30,000 Acre-ft when full, about eighteen hours worth of storage. For 5 hours per day the three reactors can provide 1.5 GW of peak power to the grid.

The Transcontinental Aqueduct. Leg 7: White Oaks Canyon Lake (to be made) to the Arch Lewis Canyon Lake via a 20 mile tunnel under the Guadaloupe Mountains in New Mexico.

Dam 1 dams the White Oaks Canyon Lake. It has a 2000 feet wide and up to 400 feet high dam, topping out at 5000 feet, and the lake holds a volume of up to 80,000 acre-ft of water.

Dam 2 dams the Last Chance Canyon Lake. It has a 2200 feet wide and up to 380 feet high dam, topping out at 5680 feet, and the lake holds a water volume of up to 35,000 acre-ft.

The Stage 7 is a tunnel, starting at 4600 feet and ending at 4492 feet maximum levels. The 20 mile long tunnel will drop 88 feet as it passes under the mountain.

Dam 3 dams the Upper Canyon Lake for lack of a better word. It has a 1600 feet wide and up to 240 feet high dam, topping out at 5200 feet, and the lake holds a volume of up to 15,000 acre-ft of water.

Dam 4 dams the Arch Lewis Canyon Lake for lack of a better word. It has a 3000 feet wide and up to 480 feet high dam, topping out at 4600 feet, and the lake holds a volume of up to 60,000 acre-ft of water.

Up to now all stages have pumped water up the mountains. This stage releases the hydroelectric water storage, and it does so even during peak power, so the water flows all 24 hours with peak electricity creation during peak usage. By now, the average flow is down to 19000 cfs , 24 hours a day. During off peak hours, 19000 cfs flows down the tunnel, the power generated is coming from Dam 1 with a water level of between 4980 feet and 4700 feet with an average of 4940 feet. The maximum output level of the water is 4640 feet, so a drop of 300 feet will generate a minimum of 450 MW of power, or 8.6 GWh/day. Part of this energy will be used to pump up the water to Dam 2 and 3. Dam 2 will pump 13,000 cfs of water from 4630 feet to between 5820 feet ans 5520 feet, (average 5760) for 19 hours, an average lift of 1,060 feet. This required a total of 6.2 GWh of energy per day , or 326 MW pf power. Dam 3 will pump 6,000 cfs of water from 4630 feet to between 5200 feet and 4930 feet, (average 5120) for 19 hours, an average lift of 520 feet. This required a total of 1.4 GWh of energy per day , or 74 MW pf power.

The net electricity generated during 19 off peak hours is 1.0 MW on average. The electricity generated during the 5 peak hours is 49,000 cfs at a drop of 1060 feet, or 1.16 GW, from dam 2. From dam 3 it will be 22,800 cfs at a drop of 520 feet, or 264 MW. From the 2 dams , total electricity is 1.42 GW. Total electricity generated during these 5 hours is 7.12 GWh. This assumes a 97% efficiency of the reversible pumps from Dam 2 and 3. The pumps from Dam 1 are not reversible.

The tunnel capacity between Dam 1 and Dam 2 outlets is 19,000 cfs, between Dam 2 and 3 it is 49,000 cfs, and from Dam 3 to its exit in Dam 4 it is 71,800 cfs.