Saturday, March 23, 2013

The Merrimack River - Renewable Energy

We all know that hydroelectric power is renewable energy. As an engineer of a sort and historian of a different sort, I've been asked a few times what the actual hydroelectric power potential of the Merrimack River at Lowell is. Many months ago I did a post that covered some of this. However, I don't quite trust my math and that wasn't the pure focus of that post, so I decided to do another one.

The questions are:

  1. How much power - at peak - can the Merrimack theoretically provide? Because of large seasonal variations in flow, this number can vary by a magnitude. So, we are also interested in how much power, in dry spells, it can reliably provide.
  2. How much water can go through the canal system? This number I figure should be very close to the historical figures. We've all seen in the summer that the river underneath the Pawtucket Falls Bridge is bone dry as it is all diverted into the canals. In the spring, there's a ton of water down there that is not being used for any mechanical purposes. 
  3. How much electricity does Lowell currently produce off of the Merrimack, at various sites around the city?
  4. Are there ways to produce more power we haven't implemented?
The first figures aren't hard to get. We just need to go to the USGS website and look at some flow figures for Lowell, since waterpower is a simple formula of volume times drop times weight of the water.

Total Power Available

So, we'll start here. This is the statistics for the amount of water in the Merrimack River below the Concord River. So, this isn't really at the point we are interested in, but for our purposes, it should be fine. You can surf around the site if you'd like, but it seems like the Concord River contributes maybe 10% of the water volume at this point, so we'll be within reason to use these figures and maybe round down as appropriate.

Two numbers in this chart are the most important: High flow and low flow.

Low flow, in August, is about 2,500 cfs (cubic feet per second). This is the reliable amount of energy available to the city. We can assume that due to the dam and lock system, the height of the canals, that is, the drop part of the equation, is constant.

High flow, in April, is nearly ten times that: 22,000 cfs. Again, the canal and dam system controls the level of the water in the system.

Now, that formula. I'm going to mention Patrick Malone's excellent Waterpower in Lowell book, which, in addition to having extensive data on the water the mills used (we'll get back to that), and the way it was captured, utilized, and researched, contains the formula I mentioned above. It also mentions, as we are often told, that the Merrimack River drops about 30 feet at Lowell and water weighs 62.4 lbs per cubic foot.

So, on the low end, we're talking about

2,500 cfs * 62.4 lbs per cf * 30 ft = 4,680,000 lb-ft per second.

Now, let's step back to high school physics. What is a pound-foot per second?

Well, a foot-pound is a unit of energy, equal to 1.356 Joules. A Joule per second is a Watt, a measure of power. So...let's convert foot-pounds to joules:

4,680,000 lb-ft * 1.356 = 6,345,228 joules

Since we're interested in joules per second, the number is directly equivalent to a watt. Our math says that at lowest flow, Lowell can expect 6.35 MW of power from the river before we factor in the energy efficiency of a turbine (which can be as high as 90%). That's a lot of lightbulbs! A horsepower is 746 watts so we're talking 8,509 horsepower, or a few dozen cars. Let that sink in as you think about gasoline: If you drive a 250hp car, your car produces about 1/30 the guaranteed power available from the entire Merrimack River at one of the best sites for an industrial city in the nation.

Now, sometimes the drop in the river is listed as 32 or even 34 feet. What's that do to these numbers? 34 feet is 7.19 MW.

Now, high flow:

22,000 cfs * 62.4 lbs per cf * 30 ft * 1.356 = 55.85 MW. Much better!

How much power is this?

Well...the US Energy Information Administration says that the average Massachusetts house uses 618 kilowatt hours a month. Let's assume that's ballpark for Lowell. Note that electricity usage is low in the Northeast because of the lack of electric heat and heavy air conditioning use. If you're using 618 kw/h a month, you're using 20.6 kw/h a day, or about 0.85 kw/h an hour. The hours cancel out so that's an instantaneous draw of 0.85 KW, or 850 Watts. You could run 850 homes on a Megawatt then, or over 5,000 homes off of the reliable energy the river can produce. Multiply by 10 for the high flow data. There are about 106,000 people in Lowell, and it looks like reasonable estimates are 30,000 households (which isn't exactly a house but close enough for us). That's one in six we could power, and that includes no other electricity uses from businesses and factories to street lights.

For comparison, the Natural Gas plant on Tanner Street, L'energia, produces 80 MW peak. The Coal/Petroleum plant on Salem Harbor is 750 MW peak. Seabrook Nuclear Power Station is 1,200 MW peak!

So, how much can the canals harness?

Again, as far as we've gotten is that the low flow for the river is 2,500 CFS. We know all of that water can get into the canal system. We know when there is a lot more than that, it flows over the top of the dam unused. We don't know how much the system can actually produce. The best way to find out the answer to this question is to look at historical data: back to Malone's book.

The thing to remember here is that in early Lowell, water was a utility. James B Francis, the most famous of the Locks and Canals engineers, the "Chief of Police of Water", really had to know how much water he had and could provide each mill so he could provide it to them reliably - and charge them for it. There was a standard unit for sale called the Mill Power, which was equivalent to 25 cubic feet per second of water falling 30 feet, or 85 theoretical horsepower. Malone has all this data. At the build-out of the canal system, Lowell produced a reliable 140 "Permanent" Mill Powers, which would be 11,900 horsepower, or 15.95 MW. This is three times the expected minimum (but under a third of spring flow), which is why the dam systems were needed to keep the water levels high enough during the day - we turn the river into a giant mill pond that refills overnight if need be. Mill companies were allowed to buy extra water when it was available, and in 1859 (after the system was completed and before heavy use of steam power), Lowell was often using an extra 40 Mill Powers. Let's call that an easy extra six Megawatts for 22 MW. There is complex engineering related to flow rates lowering head and becoming counterproductive, plus backed-up water slowing wheels when the river was high - for the book.

How much is installed?

Well, Locks and Canals, the company that provided the water, eventually ended up being owned by Italian energy conglomerate Enel. Enel, as Boott Hydroelectric, has a few sites in Lowell's mills as well as the main plant on the Northern Canal by the University Ave Bridge. Their statistics are surprisingly missing from the EIA map above, but they provide it here.

Their main plant uses 3300 cfs of water and drops it 37 feet. Using our formula to check their math, we get 10.33 MW. This is considerably lower than the 17.3 MW installed they claim here and there is no explanation for the gap. However, adding in the supplemental plants in the old mills, we get to 20.7 MW installed. If we take Lowell's historic max with surplus mill powers of 22 MW and we multiply by an efficiency factor of 90%, we get 19.8 MW... which is essentially what Enel says they can produce at maximum.

What I'm seeing is that if they are producing this power regularly, we are using all of the available water Lowell historically has had and this is not enough electricity, even in high water scenarios, to run the residential sector of the city.

Is there more somewhere?

An interesting fact about hydro power is that the technologies, because turbines are so efficient, haven't changed much since Francis was doing his experiments 150 years ago. Is there anywhere else we can squeeze out any more power at any price? Raising the dam would work, but we would flood out our neighbors upriver. There probably isn't much extra summer water, as we're already in control of the water all the way up the Merrimack's watershed to its source at Winnipesaukee.

The only way to get power out of water is to utilize its kinetic energy - it has to be moving. The Wikipedia article on hydraulic head provides a good graphic showing how a turbine, like those in Lowell, work. If you've ever seen a boat on the Merrimack in Tyngsboro or up by the Rourke Bridge up to the buoys above the dam, you know there isn't much motion to that water - it's being backed up by the dam. Go too much downstream, and you hit the 17 MW Lawrence dam installation - again, the water is not moving much above this point.

However, you have to wonder if during the summer months if it would be cost effective to install small run-of-the-river installations to catch the more-than-half of the power potential the river has. 

However, I'm not going to pretend to know anything about the math on that ;-)

In conclusion...

It would appear we are using the majority of water that Lowell has ever had available to it for power generation purposes already, and, while not an insignificant amount, is not anywhere near the electricity needs of the modern city...never mind its resource-poor suburbs. Always, always be conscious about claims of renewables replacing our dwindling fossil fuel reserves.


  1. An essentially accurate summation. Inconsistencies regarding Boot Mills, Boot Hydro-power can be fixed.

    Using the low-water flows is a better place to start, but ignores the Northern Canal and the flow rate(s) through the gates there where the head is greatest and consistent.

    It is not doing any good to mention the average Lowell home usage or any number of homes that might be powered by any source. Of particular interest should not be the commercial viability or profitability of any hydro-electric; and renewable, generation system. What should be factored and used would be the sum total of the public buildings and hospitals as well as street signal and lighting.

    Using those numbers would better poise the argument for the cost-benefit to the City of Lowell and garner the greatest public support since all of this would move to reduce our tax burden by greatly reducing or eliminating the costs associated with electric power.

    I can envision a partnership with UMass Lowell, with a herd of qualified staff and a pool of several hundred annual interns as a workforce to follow in the footsteps of Mr. Francis.

    I also envision tapping the Concord River at the confluence to double-down on our kinetic resource.

    Serving the public and not National Grid or Enel.

    Jim Buba
    Lowell, MA

  2. Jim - what do you mean that the low-water flows ignores the Northern canal? The water in that canal passes through the gatehouse on the side of the Pawtucket Falls bridge. Like the water that is passing through Guard Locks a bit upstream, it's drawn from behind the dam.

    I too would like to see the City own our water power - my understanding is right now that electricity is sold to the South Shore.

    The Concord River is capable of producing some power, and in a run-of-the-river setup (no way we could build a mill pond), we might get something. However, the USGS data shows that we get 250 cfs to 1500 cfs of flow below where River Meadow Brook enters. I don't have the elevation drop from say, Lawrence Street to East Merrimack Street in front of me but I believe it is on the same 30 foot magnitude as the Merrimack. This is water in the kilowatt range.

  3. Our neighbors north of the falls have not only been flooded but pay thousands in rising flood insurance rates and more homes are considered to be in the floodplain while the city continues to grant building permits. Lowell paid $20 M for the 2006 flood damage, Enel paid nothing though they had fortified the flashboards and pins.
    Just think of a partnership with UML where students could get hands-on experience in hydropower. Just think of Lowell once again benefiting from power generated here instead of sending it somewhere else.
    For generations the Pawtucket dam made profits for its owners and failed seasonally to prevent flooding. It works simply and simply works. The dam should once again be owned by Lowell.

  4. Anonymous - what changed the floodplain to our north by the way? Downtown, it's been bumbling by FEMA since we are on the canals, yet in nearly 200 years have never flooded. And yeah, I don't know why so much building was ever allowed in the 100-year-flood areas or why it continues. I wonder how much extra water being held back stronger pins and different boards translates to. Has there been an engineering study?

    I've said very little about the flooding issues and the dam because I feel I simply don't have enough facts. 2006 and 2007 were top-five-ever events that flooded a lot of homes that didn't exist in 1936 and 1938. The only comparable event I'm aware of in at all recent history is the April 1987 flood. I was awful young then and living on high ground in Tyngsboro, but The Boston Globe says there was standing water on the streets in downtown Lawrence. These five events are the same major five upriver in Manchester and downriver in Lawrence...

  5. This is the first time I am visiting your blog. I found so many interesting stuff in your blog especially its discussion. From the tons of comments on your articles, I guess I am not the only one having all the enjoyment here!
    Recycling of Waste in Bangalore

  6. I don't even know how I ended up here, but I thought
    this post was great. I don't know who you are but certainly you're going to a famous blogger if you
    are not already ;) Cheers!

    Also visit my blog post Ricky Salvador

  7. From October 2007 to June 2015 The Merrimack River below the Nashua River averaged 3,300 cfs. During that same time frame the river averaged about 2.5 to 3 months above 10,000 cfs. You would be lucky to get a month each year over 20,000 cfs; however, that would be irrelevant, since that kind of flow would bring about a substantial increase in a plants tailrace, with a reduction in power generation output. Even 10,000 cfs would have an impact, but it would be minimal. Enel claims that the net head for the facility is 37 ft. No reason to doubt that. By eye, it looks close to 40' which is similar to Cataract project on the Saco River in Saco/Biddeford, Maine. Therefore, with an average annual flow of 3,300 cfs and a net head of about 37 ft., you are looking at: 3,300 * 62.4 * 37 / 550 * .83 * .000746 = 8.6 mw's.
    The .83 is the general average % in efficiency for hydroelectric generators. The best I have seen in hydro power generation is the brand new General Electric stators with new digital technology Voith Kaplan Turbines put into service at the Skelton project last year in Maine. Those two 10 mw units were getting 86% efficiency. Anything above that is unheard of. Only natural gas combined cycle steam turbines will do better than 86%. The 550 is the conversion number for mechanical energy.
    During the 2 to 3 months a year that the Merrimack River is passing 10,000 cfs, the hydro plant in Lowell could potentially produce about 24 mw’s. (10,000 * 62.4 * 34 / 550 * .83 * .000746 = ~24 mw’s. The reason I say potentially is because I am not sure how much the Station’s tailrace will come up and impact the net head at 10,000 cfs. I’m guessing it would come up about 3 ft. changing the net head from 37 ft. to 34 ft. Also with high flows comes lots of debris, which will clog up and block off the generator intake racks. To add more generation to what they have there now would take a good number of years, (running 2 to 3 months a year) for the equipment to pay for itself.

    1. Thanks for the clarifications. Sounds like you do this professionally?