Our outlook on summer reservoir levels

While summer doesn’t officially start until June 21, to most of us in the Savannah River Basin, it feels like summer is already here. Temperatures are heating up, the ground is drying, and the trees are lush and green as they enter their peak season.

So what does that mean for reservoir levels? As the summer progresses, we will likely experience a dip in lake levels.

“Even with 100 percent of normal rainfall, the pools rise and fall at certain times of the year,” said Stan Simpson, a hydrologist with the Savannah District. “This is evident in the pool plots when you look at the average elevation line.”

Simpson said normal inflow in the summer is typically insufficient to hold pools at a steady elevation, but the opposite occurs in the winter and spring—when conditions are typically cooler, the ground is more saturated, and evaporation rates are lower. During those periods, the reservoirs normally rise with normal rainfall.

“A point to keep in mind is that rainfall does not vary significantly throughout the year—therefore other environmental considerations such as temperature, evaporation, and transpiration play a larger role in pool elevations,” Simpson said.

In the Savannah River Basin, evaporation alone accounts for up to 1,200 cubic feet of water loss – every second. Transpiration is much more difficult to measure, but in the summer more water is lost from transpiration than evaporation. According to the U.S. Geological Survey website, a large oak tree can transpire 40,000 gallons per year. The most tangible way these two phenomena can be recognized is observation of lake-level trends in the summer. Despite significant amounts of above-average rain in the summer, this warmer season is almost always associated with falling lake levels.

Additionally, human needs vary throughout the year. The coldest months and the hottest months tend to increase demand for electricity, which results in greater hydropower generation through increased releases.

When operating in normal conditions, our protocols include being responsive to the hydropower needs of the Southeastern Power Administration, who is under contract with their customers. According to our water control manual, the reservoirs are in normal conditions during the summer when water levels are within the first four feet of conservation storage.

Check the latest weekly projections every Wednesday by visiting our water management website at http://water.sas.usace.army.mil  and click the “Declarations” button on the left menu.

As always, we welcome your questions and comments in the comments section below. Thanks for reading us!

~Tracy Robillard, public affairs specialist

 

About US Army Corps of Engineers Savannah District

The U.S. Army Corps of Engineers Savannah District oversees a multi-million dollar military construction program at 11 Army and Air Force installations in Georgia and North Carolina. We also manage water resources across the Coastal Georgia region, including maintenance dredging of the Savannah and Brunswick harbors; operation of three hydroelectric dams and reservoirs along the upper Savannah River; and administration of an extensive stream and wetland permitting and mitigation program within the state of Georgia. Follow us on Twitter @SavannahCorps and on Facebook.com/SavannahCorps
This entry was posted in Water Management and tagged , , , , , , , , , , , , , . Bookmark the permalink.
  • Bert P Ellis

    Great article… and Stan is always on mark.

  • PBYC

    FYI 1,200 cubic feet = 8,976 gallons

  • Ferris

    “Rainfall does not vary significantly throughout the year” is inconsistent with monthly rainfall averages and discounts the affect higher winter rainfall has on the need for a winter drawdown, potentially slanting the current study considering reducing the drawdown to just 2 feet. Average monthly rainfall for the three reservoir collection area is 5.0″ for Jan, Feb, and Mar (JFM), which is 26% greater than the 3.9″ average for Sep, Oct, and Nov (SON) and 12% greater than the 4.5″ average for Jun, Jul, and Aug (JJA). A 26% or 12% increase in income is significant for most of us, but does it significantly affect the reservoir levels and basin release rate?

    As a process engineer with a specialty in system controls, I have often found myself in the situation of having to prepare an order of magnitude system performance estimate prior to budget approval and final design. This summary comes from that background experience. The total collection basin for the three reservoirs is 3.9 million acres. Assume that 55% of the JFM land rainfall plus 100% of the reservoir rainfall make it to the reservoirs, yielding a 3 month total of 2.8 million Acre-Feet (AF) of water. Hartwell and Thurmond absorb 0.5 million AF as their levels increase from 4 feet below full pool to full pool. The remaining 2.3 million AF is released over three months at an average rate of 12.6K CFS, after allowing for a monthly 200 CFS evaporation rate. Now assume that 20% of the SON land rainfall plus 100% of the reservoir rainfall make it to the reservoirs, yielding a 3 month total of 0.9 million AF of water. The release rate would average 4.2 K CFS at a constant reservoir level after allowing for a monthly 800 CFS evaporation rate. The assumptions were adjusted to be consistent with historical information and could be improved with more time and data. This summary illustrates that the ecosystem probably has more than double the effect on rainfall reaching the reservoirs compared to a rainfall difference of 26%. However, if the SON rainfall fell in JFM, the average release rate without refilling the reservoirs would be 12.0K CFS after allowing for a monthly 200 CFS evaporation rate, which is a very significant difference and might eliminate the need for a winter drawdown. Bottom line, seasonal rainfall variation is significantly important to basin management.

    Increased precipitation during JFM is the major reason Hartwell and Thurmond guide curves are 4 feet below full pool on Jan 1. In addition, the ground is saturated which allows more of the rainfall to reach the reservoirs, temperatures are cooler with reduced reservoir and land evaporation rates, and deciduous trees are without leaves with reduced transpiration rates. Rainfall variations throughout the year are significant to basin management, with higher rainfall occurring during JFM when the ecosystem is less able to absorb it and lower rainfall occurring during JJA and SON when the ecosystem is more able to absorb it. Rainfall rates and ecosystem status combine to keep the water manager drawing on experience and studying a “crystal ball” to set release rates that meet the needs of upstream and downstream users.

    ~Ferris

    • http://www.sas.usace.army.mil/ US Army Corps of Engineers

      Interesting analysis, Ferris. Comparing rainfall differences to income differences caught my attention but I’m not sure it accurately fits. My graduate school education in economics piques my curiosity in the comparison, but also gives me reason to be skeptical of the comparison. It would require more study.

      That being said, I paid closest attention to your final statement: “Rainfall rates and ecosystem status combine to keep the water manager drawing on experience and studying a ‘crystal ball’ to set release rates that meet the needs of upstream and downstream users.” This is a profound observation demonstrating that years of study and observation of the Savannah River basin weigh heavily on our management decisions. Our water managers don’t rely solely on computer models or theories, but on decades of working in this basin under a wide range of conditions. They’ve seen floods, droughts, heat waves, ice storms, tropical storms, and years of other weather.

      Our water managers are scientist-engineers and draw on the latest scientific methods as well as years of real-world observations to accomplish all their missions in the most efficient way possible. Although they don’t have ‘crystal balls’ to consult, they actually have something better – science tempered with experience and insight.
      ~Billy Birdwell, Corporate Communications Office

      • Ferris

        Hi Billy, thanks for the response. I very much appreciate the scientist/engineer viewpoint as a Chemical Engineer who has spent as much time in college education and work experience studying and using science as engineering. I have observed release rates being increased in anticipation of heavy rainfall events. Although anticipating the weather does require science and experience, it also requires a bit of what I refer to as “crystal ball” studying, whatever your official terminology, because there is a risk that the event may not materialize or may be greater than anticipated.

        To restate my point using monthly rainfall data, I find it incredible that USACE refers to a 28% increase in rainfall from one 3 month period to another as “not very significant”. It must certainly be considered significant that Mar rainfall is 51% greater than Oct rainfall. The increased power generation and income resulting from the release of 28% or 51% more water is surely a reasonable comparison to increased personal income for those lacking a graduate economics degree or process experience. Having that rainfall increase when the basin is less able to absorb the water amplifies the affect by 28% over the discussed periods and by 51% when comparing Mar to Oct. That should be significant in anyone’s viewpoint. We would be in a heap of trouble if the water manager did not understand the significance. ~Ferris

        • http://www.sas.usace.army.mil/ US Army Corps of Engineers

          Hello Ferris – Billy is unavailable to respond, so I am addressing your comment on his behalf.

          One point needs to be corrected outright: We do not set release rates based on rainfall projections. Our operating protocols involve responding only to observed rainfall – that is rainfall that has already fallen in the basin. Your premise being unsound, the “crystal ball” conclusion is therefore of no value.

          Concerning your rainfall data, I’m not sure what source you are using, but our data does not show anything like a 51% variance in October and March rainfall – not in any of the sub basins or total combined. Counterpoint aside, wrestling over what should be considered “significant” rainfall is unprofitable – and not the least reason being that the term “significant” is subjective by nature and highly vulnerable to interpretation. The more important reason is that observed rainfall, by itself, is not a reliable indicator of reservoir levels. Other considerations, such as season, soil conditions, temperature, and land development affect how much rain water actually turns into runoff and makes it into the reservoirs, (which is usually less than 40% of observed rainfall). Predicting runoff in relation to rainfall, then, is the real key to making progress on projections. That is the real point Stan Simpson was getting at in his statement on rainfall variance throughout the season.

          I hope this helps ~Russell Wicke

          • Ferris

            I would have to relocate the link, but my source is from tables printed in March of this year from your site that show average rainfall in the three basins from 1948. Thurmond, for example, shows an average rainfall of 3.01″ in October and an average rainfall of 4.75″ in March, or 57.8% more rainfall on average in March than October. Blogs posted by Russell and Tracy, under the category of Rainfall Update, list these rainfall averages as 3.0″ and 4.8″ respectively, so my numbers are right on target.

            Predicting runoff in relation to rainfall during the different conditions previously discussed is very important and I think should be a primary objective. I would expect a much higher runoff rate in March than October, amplified by the 57.8% increase in average rainfall. I fail to understand how anyone could consider that insignificant. X * 1.578 = 1.578X

            I am interested in what is correct rather than who is correct. By accounts, Stan is a great guy liked by many people; but his statement that rainfall does not vary significantly throughout the year is simply not correct. An lower than needed drawdown will result if that philosophy is used to determine the amount required for the winter drawdown, and that is why I am so interested.

            Does USACE ever admit they are wrong? Blindly defending a position severely limits opportunities for improvement.
            ~Ferris

          • http://www.sas.usace.army.mil/ US Army Corps of Engineers

            Ferris – this is an example of why the percentage narrative does not profit us in our circumstance: I can take the same numbers and say it is only a 36.6% decrease from 4.75 to 3.01. We’re using the same data and taking the numbers in different directions. It’s not helpful. Furthermore, as I suggested above, what a person terms as “insignificant” is highly subjective and completely depends on the context within which it is being measured. The statement, isolated from hard data, qualifying facts and context, cannot be said to be correct or incorrect. With the information you have you are defending an impossible assertion because it lacks a perspective framework.

            For example, compared to observed rainfall, average rainfall truly doesn’t vary much. Consider that within the last year, Hartwell received 13.5 inches in July 2013, but only 1.6 inches in October 2013. And this past March – the month with the highest average – Hartwell only recorded 3.9 inches compared to July’s 13.5 inches. These are significant variances compared to average rainfall data for the same months.

            But again – none of this is helpful because your objection has little to do with the thesis of the post, which is that pools rise and fall naturally throughout the year, *even* with normal rainfall (i.e. rainfall that doesn’t vary much compared abnormally wide variances). Incidentally, this is not incompatible with the point I think you are making; which, to quote from your original post: “Bottom line, seasonal rainfall variation is significantly important to basin management.” No one here is disputing that. In fact, we’ve argued that point before. What I’m disputing is your rationale in claiming Stan’s statement is, without qualification, definitively wrong.

            ~Russell

          • Ferris

            Maybe Stan should qualify his statement?

          • http://www.sas.usace.army.mil/ US Army Corps of Engineers

            As I implied in my comment, the context of the article qualifies Stan’s statement to some degree. And I hope that my comment has assisted to qualify it to the degree necessary. ~Russell

          • Ferris

            Russell, I get it. When employees of the USACE, like you and Stan, say something is significant or not significant it is indeed a fact; however, when I use those words it is not profitable or helpful and it is highly subjective. I should disregard my decades of experience that required constantly evaluating whether something was significant regarding a plant construction budget, process variables like flow rates/temperature/pressure/level, equipment sizes, instrument measurements, scheduling, and many other details in industries as diverse as food and beverage to waste water treatment, specialty chemicals to pharmaceuticals, and wire drawing to silicon chip manufacturing. Not! ~Ferris

          • Ferris

            Russell, history shows that on Friday, Dec 20, 2013, the Thurmond release rate jumped from 7,684 CFS to 12,951 CFS. The most recent rainfall had been about 1″ six days earlier over the whole basin above the Thurmond dam. The release rate for Hartwell dropped on Dec 20 to 9,864 CFS with no releases for Sat and Sun as normal. Heavy rainfall of over 3″ was forecast over the next several days, materializing on Sun and Mon, Dec 22 & 23, with about 3″ in all three basins. The release rate at Thurmond was not increased further for eight days while the Hartwell release rate remained less than 10,000 CFS. The driving force for increasing the release rate on Dec 20 would have been the forecast rainfall on top of the already excessively high reservoir levels at Hartwell and Thurmond.

            If the desire were simply to lower the reservoir levels to the Guide Curves, or by 4.23 feet and 3.94 feet respectively, the increase in release rates should have begun Nov 27 when Hartwell and Thurmond were 2.02 feet and 0.90 feet respectively above the Guide Curves and much more manageable. However, the release rate at Hartwell remained less than 5,300 CFS through Dec 10 and the release rate at Thurmond remained less than 5,000 CFS through Dec 13. Meanwhile reservoir levels continued to rise so that Hartwell and Thurmond were 4.15 feet and 2.82 feet respectively above the Guide Curves on the days mentioned.

          • http://www.sas.usace.army.mil/ US Army Corps of Engineers

            Ferris – you are making an assumption that releases increased at that time in anticipation of the forecast. In fact, there are other factors that go into release rates, especailly during cold winter days and hot summer days. Since hydropower is used as peaking power, the plants are subject to spikes in demand providing we are not in drought conditions. Dec. 29 and 31 also showed spikes in releases, and again on Jan. 1, without the forecast of large amounts of rain. ~Russell

          • Ferris

            Russell- We must be looking at different data; the spike in release rates occurred on Dec 28, not Dec 29. The rainfall came on Dec 29. The Thurmond release rate increased from 12,573 CFS on Dec 27 to 15,294 CFS on Dec 28, just ahead of forecast heavy rain that resulted on Dec 29 in 2.18″, 1.87″, and 1.11″ of rainfall in the Thurmond, Russell, and Hartwell basins respectively. On Dec 28, Thurmond was 6.00 feet above the Guide Curve at 332.00 feet. Hartwell and Russell were also in flood storage. On Dec 31, all three reservoirs remained in flood storage with Thurmond at 333.11 feet, Russell at 476.49 feet, and Hartwell at 661.61 feet. On Jan 1, 2014, the Thurmond release rate jumped to 24,573 CFS, beginning the journey to relieve water issues upstream but adding to water issues downstream. If the 24,573 CFS was for peak demand as you are suggesting, we have a lot more to discuss. It think it is short-sighted to not consider the rainfall forecast when the reservoirs are in Flood Storage, and it certainly appears that the head water manager did wisely considered the forecast. ~Ferris

          • http://www.sas.usace.army.mil/ US Army Corps of Engineers

            Ferris – the overall point in all this data is being obscured – and I’m not even quite sure myself anymore what your objection is. ~Russell

          • Ferris

            Russell- Really? Allow me to recap.

            I wrote that “I have observed release rates being increased in anticipation of heavy rainfall events.”

            You objected to my statement by writing that “One point needs to be corrected outright: We do not set release rates based on rainfall projections. Our operating protocols involve responding only to observed rainfall – that is rainfall that has already fallen in the basin. Your premise being unsound, the “crystal ball” conclusion is therefore of no value.”

            I then provided data to support my conclusion.

            You then provided examples that that supposedly contradicted my conclusion.

            I then provided the most recent data with related details for your examples demonstrating that they either support or do not contradict my conclusion that forecast heavy rainfall events are considered when determining release rates, especially when pool levels are above the Guide Curves.

            My concern, stated twice before on this post, is that discounting the seasonal rainfall variation unfairly slants the argument in favor of reducing the winter drawdown to 2 feet from the current 4 feet. The data were provided to demonstrate the reason for my concern. My concern is also informed by the failure to closely follow the guide curves this past fall for Hartwell and Thurmond that call for 4 foot winter drawdown.

            But rather than address my concern, Billy chose to trivialize the discussion by challenging whether it is valid to compare rainfall differences to income differences, and you chose to challenge whether my statement that a 57.8% increase, or the 36.6% decrease you used, is significant. You called it “unprofitable”, “subjective”, and “not helpful”.

            You wrote that the thesis of this post is that pools rise and fall naturally throughout the year and that my objection has little to do with that. In fact, my objection is related to how the USACE response to the natural fluctuations influences the amount that pools rise and fall during the winter drawdown and following spring rains. And I do see room for improvement.

            ~Ferris

          • http://www.sas.usace.army.mil/ US Army Corps of Engineers

            Thank you for recapping Ferris. I can see I have contributed to the rise in temperature of this discussion. This wasn’t my intent, so please allow me an attempt at remediation.

            First, I will address your concern directly: The current winter draw down of four feet is not something the Corps can change at will based on preliminary or incomplete data. A change of that nature must come from Congress, and can only be initiated by the Corps – and even then only after an adequate study has demonstrated the risks to be negligible. Based on repeated feedback from our stakeholders, many of which hypothesize there is enough flood storage allocation to reduce the draw down to two feet, we initiated one such study by commissioning a scientific team outside the district (The Hydraulic Engineering Center). We refer to this study as the Flood Storage Reallocation Study. The study is extensive: it is conducted by highly experienced hydraulic engineers, costs approximately $240,000 and will about a year to complete. We gave details about this study in a post where we cautioned people that we are subject to the results of the study, which could recommend any number of actions, from reducing flood storage, to increasing it or no action at all. You can read that post here: .

            Second: I can assure you, despite the appearances, our protocols do not include releasing water based on projected rainfall. This practice would come with a high potential to waste water if the rainfall doesn’t materialize. And based on our basin’s history with drought, wasting water is unacceptable to nearly all stakeholders.

            Third: The history of last winter’s levels not remaining on point with guide curve is something that is probably best left to a discussion between you and Stan. You have stated your credentials – and I have no doubt Stan can assist you in understanding what took place at that time. My attempts have proved inadequate. But I will only state that although I have no doubt of your proficiency in the discipline of the sciences, I did not see that your specific discipline included hydrology. I am not a hydrologist either, which accounts for my limitations in explaining the situation to your satisfaction. I only know enough to recognize there is much more that goes into (and went into) calculations for outflows in general, and outflows this last winter, respectively. For example, you mentioned releases spiked at Thurmond right before a day of heavy rainfall on Dec. 20 and 28 (you caught my Dec. “29” typo). But if you query the database to show net inflows you will see in the days leading up to Dec. 20, a rapid increase in inflows, (likely a delayed response to previous rainfall on Dec. 14 – but this is a subject for Stan). Again, on Dec. 23 Thurmond got two inches of rain and inflows spike to 53,598 and held into the 50K cfs range into the next day. Inflows remained high in the following days and then we see an increase in outflows again on Dec. 28 – coincidentally right before another rain event. But let me restate: I am not the hydrologist and I didn’t direct these actions – Stan did, and you will need to talk to him. The point I’m trying to make is that there are a great many fine and nuanced details that go into determining what the release rates are for a given day, and even a given hour. If you would like to look that deeply into the process, Stan is a very good man, and I am sure he would be happy to indulge you. This blog, however, has its limitations when it comes to plying down into multiple layers of scientific data.

            You have my email address and my phone number. If you would like to speak with Stan, I will let him know and pass along his information to you.

            I truly hope this helps, and hope you will continue to read and provide feedback on this blog.
            ~Russell

          • Ferris

            Hi Russell, thanks for your detailed response to my concerns with the helpful information. I do not have hydrology experience and am eager to learn more about the thought process that goes into the decision making. ~Ferris

          • Ferris

            These items relevent to this discussion were covered by Stan and Ferris during a June 16, 2014, phone call. Russell was also a party to the call. Stan, thank you for discussing my concerns. Russell, thank you for facilitating the call. Text in braces {} are my comments.

            - The water manager has discretion within a couple of feet for the winter drawdown. Some political influence was applied to keep Hartwell about 2 feet above the winter drawdown curve. The plan was for Thurmond to drop about 2 feet below the winter drawdown curve in compensation, {I’m guessing on the rising curve in January}. The rainfall events occurred before the Thurmond portion of the plan was ready for execution {resulting in upstream and downstream water issues}.

            - Forecasts for rainfall events within a couple of days are considered when setting release rates if justified by circumstances, even though not strictly by protocol. {The water manager has considerable discretion in setting release rates.}

            - The statement “rainfall does not vary significantly throughout the year” is a comparison to other areas of the country, for example, areas with snowfall accumulation. For some areas the difference is double {the heavy period is 200% of the light period}. For our area, the increase is 51% {the heavy period is 151% of the light period}.

            - The Flood Storage Reallocation Study will evaluate about 75 years of inflow data for the winter drawdown recommendation. {Using inflow rather than rainfall alleviates my concern about the rainfall statement not reflecting how rainfall variation throughout the year affects the rise and fall of lakes in our basin.}

            - Having the river level rise into the flood zones at Burton’s Ferry and Clyo is not as much of a concern as exceeding 30K CFS at Butler Creek because of the lack of structures.

      • Ferris

        The weighted average rainfall is calculated by multiplying the specific basin rainfall times by the area of that specific basin, adding the three numbers, and dividing by the total area. The weighted average rainfall for Jan Feb Mar rainfall is 35% greater than the Sep Oct Nov amount and the Mar weighted average rainfall is 56% greater than the Oct amount. The area used for the total basin of Hartwell includes the area above two other dams because that water drains to Hartwell.

  • Fish1

    The week ending June 6 on Lake Hartwell shows the minimal impact of the prominently mentioned evaporation or transpiration, but the significant impact of the Corps releases of @6k cfs, resulting in Lake Hartwell going down 6″ in 4 days. Corps releases are where most of our water goes, precipitating low lake levels during low rainfall periods.
    In a number of years over the last decade, the Corps has celebrated the July 4th holiday week by releasing significant quantities of lakes water and thus lakes levels.

  • Ferris

    This post was copied from below because it includes insights to many comments in this lengthy discussion that were provided by Stan during a June 16, 2014, phone call with Ferris. Russell was also a party to the call. Stan, thank you for discussing my concerns. Russell, thank you for facilitating the call. Text in braces {} are my comments.

    - The water manager has discretion within a couple of feet for the winter drawdown. Some political influence was applied this past fall to keep Hartwell about 2 feet above the winter drawdown curve. The plan was for Thurmond to drop about 2 feet below the winter drawdown curve in compensation, {I’m guessing on the rising curve in January}. The rainfall events occurred before the Thurmond portion of the plan was ready for execution {resulting in upstream and downstream water issues}.

    - Forecasts for rainfall events within a couple of days are considered when setting release rates if justified by circumstances, even though not strictly by protocol. {The water manager has considerable discretion in setting release rates.}

    - The statement “rainfall does not vary significantly throughout the year” is a comparison to other areas of the country, for example, areas with snowfall accumulation. For some areas the difference is double {the heavy period is 200% of the light period}. For our area, the increase is 51% {the heavy period is 151% of the light period}.

    - The Flood Storage Reallocation Study will evaluate about 75 years of inflow data for the winter drawdown recommendation. {Using inflow rather than rainfall alleviates my concern about the rainfall statement not reflecting how rainfall variation throughout the year affects the rise and fall of lakes in our basin.}

    - Having the river level rise into the flood zones at Burton’s Ferry and Clyo is not as much of a concern as exceeding 30K CFS at Butler Creek because of the lack of structures.

    ~Ferris