turn heat into electricity

[Rematog]

Cogen for low level heat (HVAC, etc.) could be possible for distributed units. But that will be a good decade or two after initial deployment. The fact that this is the nuclear reactor (yes, I know it’s fusion…but it still will be licensed by the NRC and puts out radiation (yes, I know it’s x-ray wavelength electromagnetic)…I just do not believe you would get a permit to site one anywhere NEAR a residential or commercial area. And make no mistake, it will take a permit to install a fusion power block for the foreseeable future. (I am currently working on a project with an electrically powered x-ray tube used for analysis of coal. We will have to have a permit from the state to install it.)

Speaking of Heat, I neglected to include cooling water and plant auxiliary loads in my calculations.

I went back to my calculations, and used the inefficiencies of beam and x-ray conversion (90% eff, 80% x-ray) and calculated the waste heat. I added to this a reactor electrode cooling load of 5% of the gross energy (input + fusion) and come up with a heat load of 13.2 MMBtu per 5 MW power block or 1,586 MMBtu for a 600 MW plant re-powering with 120 FF power blocks. For comparison, from a modern 600 MW steam turbine heat balance, the condenser heat rejection is 2,492 MMBtu.

So, a cooling tower of 65% of the size can be used to cool Focus Fusion Block generating the same typical 600MW net. This works well for repowering, as the existing circulating water system would be re-used to supply cooling water.

I completely agree that my waste heat estimate is only a placeholder (esp. the electrode cooling) and better data on cooling loads is needed.

But, in all my cost estimate cases, I need to add the costs for plumbing the cooling systems up. While this adds to Focus Fusions costs, this will only decrease it’s advantage slightly.

Also, need to look at power requirements for plant site auxiliaries, including cooling tower fans and pumps, lighting, offices, etc. These will reduce plant net output by a very small percentage. For reference, a cooling tower of this size that I have data for, has 19 fans @150 Hp each and two circulating water pumps @ 3,500 Hp ea. So cooling alone will take 5.5 MW assuming 95% motor efficiency. I will re-run the busbar cost estimates with a total 3% parasitic load.

[Rematog]

Ran busbar estimate again. Didn

[Brian H]

Rematog wrote: My estimate. I am very sorry about the readability. I’ve yet to find a way to get this forum to format tables or allow posting files.

Rematog

Power Block Fabrication
Controls $10,000 PLC with local screen
Building $20,000 Prefab building, skid mounted
Cables, conduit, etc $20,000 power and control
MCC, Low voltage xformer $20,000 power distribution, 480V aux power
Solid State Power Controls $50,000 specialized high speed switching (to pulse reactor)
Capacitors, power supply $50,000 SWAG
HV xformer, disconnects $20,000 Output step up, required isolation devices.
Reactor Vessel $5,000 SWAG
Discharge Electrodes $5,000 SWAG
Vacuum pump $5,000 SWAG
Fuel system $5,000 SWAG
Ion beam power converter $10,000 SWAG
X-ray power converter $20,000 SWAG
Shielding $20,000 SWAG
Radiation detectors, other sensors $10,000 SWAG
Fabrication Labor $36,000 1200 $30.00 30 man-weeks @ 40 hr/wk
Overhead and Profit $100,000 Very arguable number
–
$406,000
Site Work, Installation
Foundation $20,000 excavation, fill, forms, rebar and mud, incs labor
Site Work (civil, fences, roads) $5,000 $500,000 Additions and changes
Control Room, etc $50,000 $5,000,000 Oversight for all system, DCS and Datalogging
Installation Labor $50,000 1000 $50.00 20 man-weeks @ 50 hr/wk (typ field work)
Shipping $5,000 for Power Block
Interconnection Cables $20,000 Power to transmission transformer yard
Engineering, supervision $10,000 $1,000,000
Rental Equipment $10,000
–
$170,000

Site Cost, New Development
Land $5,000 1/2 acre @ $10k/acre
Site Work (civil, fences, roads) $20,000 $2,000,000
Other Site Buildings $25,000 $2,500,000 Offices, maint, warehouse
Transformer Yard $200,000 $20,000,000
–
$250,000

Brown Field Cost $576,000
Green Field Cost $826,000

Lots of items there that don’t quite mesh with available info.

1) the units are conceived as factory pre-fab, trucked in and erected. The man-hours for installation etc. estimate is way high, probably at least an order of magnitude.
2) capacitors are off-the-shelf, and $50,000 seems high
3) profit for the installed unit seems questionable, as the revenue projections use licensing as pretty much the sole profit source.
4) Normal grid interfaces are not necessarily a good model for connection costs, as distributed generation would argue against large “brown field” clusters as the dominant model. Many, particularly in industrial zones, would likely be primarily dedicated to local consumption. Further, standardized interfaces would certainly be developed and implemented at far less than these projected costs for the volumes of installations projected.
5) Offices & site buildings are overstated, IMO; the projection I saw was for one office housing one or two technicians per 5 or 10 generation sites, remotely monitored.
6) Land costs are not part of any estimate process for the units. They would vary from $0 to a few hundred thousand perhaps, for the garage-sized building and interconnect facility, and the control site is a “shared” cost between 5 or 10, and might even be an office rented in a local commercial building.

Etc.

[Rematog]

Brian H wrote:

Lots of items there that don’t quite mesh with available info.

1) the units are conceived as factory pre-fab, trucked in and erected. The man-hours for installation etc. estimate is way high, probably at least an order of magnitude.
2) capacitors are off-the-shelf, and $50,000 seems high
3) profit for the installed unit seems questionable, as the revenue projections use licensing as pretty much the sole profit source.
4) Normal grid interfaces are not necessarily a good model for connection costs, as distributed generation would argue against large “brown field” clusters as the dominant model. Many, particularly in industrial zones, would likely be primarily dedicated to local consumption. Further, standardized interfaces would certainly be developed and implemented at far less than these projected costs for the volumes of installations projected.
5) Offices & site buildings are overstated, IMO; the projection I saw was for one office housing one or two technicians per 5 or 10 generation sites, remotely monitored.
6) Land costs are not part of any estimate process for the units. They would vary from $0 to a few hundred thousand perhaps, for the garage-sized building and interconnect facility, and the control site is a “shared” cost between 5 or 10, and might even be an office rented in a local commercial building.

Etc.

Brian,

Thank you for responding by my post. I enjoy getting intelligent, thought out feedback.

Pre-Fab: I’m sorry if I wasn’t clear that the first section of the estimate is for the factory fab. of the power block. I did this as a “reality check” of the number given by Focus Fusion. I intentionally kept cost what I considered very low. So, to me, this showed their number was reasonable as a very low end estimate.

Capacitors: are indeed off the shelf. An I have no data other the capacitance needed, nor did I try pricing by contacting vendors. I used a number that seemed reasonable for something that big, operating at 3-4 Kv. If you know what a real price would be, please post it and I’ll use it.

Profit: The Overhead and Profit figure I put in ($100,000K/module) was for the company factory fabricating the modules, not Focus Fusion. The royalty is part of the overhead, which also includes cost of owning and operating a factory, insurance, taxes, management and engineering salaries, etc. I used roughly a 25% mark-up for the O&P figure. If someone has real life business experience with the O&P for large, high tech skid mounted equipment in series production, please post and I’ll use it.

Distributed Generation: I agree that distributed generation was advantages that will drive the industry in that direction in the long run. But, in my posts, I am considering more the initial phase of deployment, say the first decade or two. In that time frame, due to the fact of this being a nuclear technology (fusion is a nuclear reaction), I firmly believe that the NRC, local regulation, and especially local opinion as expressed by zoning, permits and intervention, would prohibit it’s use in urban areas. Second, several economic factors (ease of maintenance, operations/oversight, security and economies of scale in construction) would make having the power block located on larger sites attractive. Third, the organizations doing the installation would be, again in my opinion, utilities, whether investor owned or governmental, and this will have advantage to those organizations of allowing reuse of facilitates. Think of this as being the business version of “realpolitik”.

Office, buildings, Control room: Again, my estimate is based on a large site with a lot of modules. I did divide the cost by the number of modules, so per unit, it is not that high. The control room I’m discussing has a DCS system (Distributed Control System, an industrial automation system to accept analog, discreet (on/off) and digital field signals, process them, make control outputs, again analog, discreet and digital. These systems provide user interfaces via multiple touch screen stations) which can monitor and control a large, complex operation. Other more mundane, but necessary, things also fall under this and site improvements, such as roads, outdoor lighting, water and sewer, etc. While I would not say my estimates are dead on, I would argue that based on my experience, they represent the low end of the cost range.

Land: You say they are not part of a cost estimate, then go one to say this is because they are extremely variable. I agree they are variable, but that does not, to me, mean they should be ignored in a greenfield cost estimate. If you disagree with my estimated value, fine, lets discuss why. One of the advantages I mention for brownfield development is the (arguable) assumption that the land on the existing site is already paid for and thus “free”. And re-use of brownfields does reduce environmental impact, worthwhile in itself.

In general, remember that I am using the standard of a heavy industrial installation with a 30 year design life, meeting OSHA, NEC, NEMA, ASME, ASCE, HEI, NFP and other codes and standards (including the insurance carrier’s) that all industrial facilities in the United States must meet. Yes, this stuff makes it more expensive, but all existing power technologies, and the prices your comparing Focus Fusion to, meet these requirements. It would be unfair, and un-realist, to make a comparison to Focus Fusion built to light commercial standards, not meeting codes. I.e. you could not expect to put a reactor in a lightly built trailer and sting cables from the ceiling, then park it in a neighborhood.

And, thinking about it, if the NRC gets involved with design, QA/QC requirements, etc., expect the cost to go up something like ten fold. (Modules costing $3-5 million ea, at factory, uninstalled).

[Rematog]

Cooling Load and Cooling tower design.

I’ve been looking into cooling loads needed for a 5MW Focus Fusion Power Block.

My Assumptions:

Net output power is 5000kw
raw ion power is 98% of input power
raw x-ray power is 57% of input power
raw ion power coversion to electrical power eff. = 90%
raw x-ray power conversion to electrical power eff = 80%
Therefore: 0.98(Input) x 90% + 0.57(Input) x 80% = 5000kw + (Input)
so: 1.338(Input) = 5000kw + (Input)
Input = 14,793 Kw

Heat wasted is (100% – coversion eff) x raw power
Heat(Ion Beam) = 0.98(14,793kw) x 10% = 1450 kw
Heat(X-Ray) = 0.57(14793kw) x 20% = 1686 kw
Allowance for heat at electrode 5% of Input power = 14793kw x 5% = 740 kw.
Total Heat rejected 1450kw + 1686kw + 740 kw = 3876 kw = 13.2 MMBtu/hr for 5 MW net output.

Assume heat load linear with scaling reactor block output.

Cooling Tower Design

I’ve just gotten a cost estimate for a cooling tower which would be sufficient to meet the cooling needs for 200 MW of Focus Fusion output, based on my assumptions stated above.

I’d be glad to forward a PDF of the quote, and the URL of the site which provided it.

Bottom line was bare cooling tower materials $434K, FOB Factory. My rough estimate for installed cost is $625K, not including the circulating water pumps and piping, blow down systems and waste water disposal, make-up water supply and treatment and land.

Result summary: a bare Cooling tower will add $3.12/kw to the capital cost. It’s fans will consume 157kw of the Fusion blocks net output. The cooling water system would have a flow rate of 29,000 gpm and the tower would impose a 24′ head on the cooling water system (to be added to piping pressure loss and delta P flange to flange of the Focus Fusion power blocks cooling system heat exchangers). This was based on an assumed hot water temp of 140F and cool water temp of 100F. It was designed for a Louisiana environment.

The Hot temp is an important design/cost factor, as hotter water is easier to cool with a given ambient air temperature and humidity. I ran the cost estimate again with 180F hot water and 1/2 the flow rate (same heat rejected). Capital cost would be roughly 65%, fan Hp 125% (250 vs 200 HP) and cooling water pump head 2′ more (fan HP more as the same heat is rejected at higher temp, so outlet air temp higher, air less dense, more volume to move….)

In both cases I assumed the more expensive fiberglass construction as opposed to cheaper wood. FG lasts longer and requires less maintenance. For a project with a 30 yr design life, FG gets the nod.

If anyone wonders why I’m assuming a water based cooling system, just ask….

[Lerner]

I would think that the whole question of getting rid of the waste heat makes the smaller units better. At some point you don’t need a cooling tower, right? You can just dump the heat into the sewer system. Is 5 MW too big to do this, Rematog? If it is not, the capital savings would be large and would make up for the technician’s travel time between sites.

Even if you need a cooling tower, does the cost scale linearly with power?

Capital cost is also why we don’t intend to turn the heat inot electricity. Using the heat directly in urban areas for heating is a lot cheaper and might be the best way to get rid of it.

[Rematog]

The only thought I’ve had on size has been; Can you cool the electrode as output goes up due to higher “pulse rate”? I know next to nothing about the heatload anticipated at the electrode, hence my stated 5% placeholder assumption for that.

By my assumptions, the cooling water flow (using a 40F rise on the cooling water temp) is 730 gpm. This a big flow by some standards (12 gallons per second, a big hose, not quite a fire hose though).

In the US, there are temperature limits and total heat limits on flows returning to bodies of water. This is why most plant’s have a cooling tower. Note, the tower consumes a significant amount of water, both thru evaporation and due to the necessary blowdown to control solids concentration (referred to as “cycles of concentration”).

Scaling is generally linear over a limited range, but some items (electrical equipment costs come to mind) don’t change as fast as capacity. Economy of scale does apply.

Just to give an idea of scale, the pipes for the cooling towers where I work are 8 foot inside diameter. A rough conversion estimate is that this would be the size needed to support 1600MW of FF power blocks (The pipe serves a 575MW steam unit, a big part of the difference, the steam unit has a lower hot water temperature due to need to keep condenser back pressure to as low as possible in order to maximize cycle efficiency).

Reusing waste heat is done where there is a useful load for low temperature heat. District heating is popular in Northern Europe. Industrial uses happen too, both in US and elsewhere. FF may have an advantage, as it’s electrical generation efficiency and output would be mostly independent of outlet cooling water temperature (not true of steam turbines, as I mentioned above). A higher outlet cooling water temperature will expand the uses to which it can be put (Hotter water can be used for more things). It will also reduce the water flow required for a given heat load, and if cooling towers are used, reduce the size and cost.

This condenser cost reduction effect would be especially important in dry regions, where air cooling of the water (by fined tubes as opposed to direct air/water contact) is needed to eliminate the water usage for the cooling tower. Air cooling by finned tubes is more expensive, both of capital cost and in parasitic load for fans, pumps etc.

So even in area of waste heat rejection, Focus Fusion has technical/cost advantages.

[Rematog]

Quick thought on economies of scale.

Which do you think would be less expensive and lower maintenance:

One 5000 hp diesel engine or twenty 250 hp V-8’s.

Rematog

[Lerner]

Just to be clear about cooling: Cooling the anode, the inner electrode, will be an engineering challenge. The direct coolant will almost certainly be helium under high pressure. This is often used for intense radiation environments, because helium is very transparent to x-rays. You could not use water because the oxygen would sop up x-rays.

The basic challenge is getting heat out very quickly from a small area.

Once you do that, the helium can be cooled in a heat-exchanger with water using commercial technology.

Also, I did not think cooling towers are so cheap–that sounds awfully low: $15,000 for a 5MW cooling tower–is that right?

One more thing, if we did have just one 5MW unit, would that not be small enough to be air-cooled, like a car’s radiator?

[Lerner]

I just looked this up–diesel locomtive engines can be several MW and clearly convert most of their fuel into waste heat, yet they are air-cooled, that is their radiators are. Why could not a FF reactor just have an air-cooled radiator? No water flow is needed, in fact you might just have the helium in the radiator. This would be an environmental gain.

I can see you can’t just dump the heat in the sewer. In NYC just 1.5 GW of cooling would raise the water temperature to 100F.

[Rematog]

Eric,

I can understand why you would use helium for your cooling fluid for the electrode (I’m using this term for the cathode/anode combination, is this correct?). For completely different reasons, large (>100MW) generators are commonly cooled by hydrogen gas. The gas is then cooled in turn by the cooling water, with the heat exchange occurring in a shell and tube heat exchanger. This just adds another piece of capital equipment, and something else to maintain and monitor. (even low maintenance items need repair eventually).

Regarding the cost, I believe I answered the question of cost linearity with a statement saying “Generally linear within a limited range”. I had in mind something on the order of 1/2 the size costing 60% as much (or, conversly, twice the size being 180% the cost). Please note my figure of $3.12/kw for cooling tower (without cooling water pumps, piping, make-up water and blowdown systems) was for a (relatively small) 200MW size. A single 5 MW FF power block is a 40:1 size reduction from this. The cost would not be linear over that range. For a very small system, the cost would most likely be N times as much per Kw, with N being a small integer. I have not tried to price a system this tiny, so don’t really have a good feel for it.

While a direct water/air contact cooling tower (a “wet” tower) is lower in cost than a non-contact (water in finned tubes, referred to as a dry type cooling tower) at the larger sizes used in the power industry, It may well be that at the small size of a single 5 MW power block, a radiator type cooler is the most cost effective way to reject the heat from the cooling water into the atmosphere. But I would be quite sure the cost per KW will be more than $3.12 per KW I calculated for a much larger system, and the fan horsepower for moving cooling air will be greater as well (more air needed as only sensible heat added to air, no evaporative cooling with dry types such as a radiator). Again, this is an example of economies of scale.

Yes, you could use helium directly cooled by helium to air radiator, but this would mean a larger radiator. It would take an engineering study to determine if that was truly less expensive. Water’s high thermal carrying capacity, plus it’s low cost and lack of dangerous traits (with corrosion inhibitors) are why it is the most commonly used coolant. Helium will have more problems with coolant leakage than water and would cost much more to keep the system full of coolant (cost of replacement helium).

Regarding internal combustion engine (diesel or gasoline), yes, a lot of the fuel is converted to waste heat. And a good portion goes out the exhaust pipe as sensible heat in the exhaust gases and heat of evaporation of the water formed from combustion of the hydrogen part the of fuel. And another large part of the waste heat is transferred to the cooling water and rejected by the radiator. I imagine if someone was to try and put 200 large diesels on one site to generate 1GW, they would find a cooling tower less expensive then the 200 radiators, if the radiator was not already part of the standard engine package. Remember, the radiators would all have to have a source of cool air and a way to vent the heated air.

I will repeat my prior statements that I feel it highly unlikely that, at least in the US or Western Europe, distributed fusion reactors, with a fuel system using a neurotoxic, skin absorb-able gas, would be permitted at an unattended site or in urban areas, for at least a couple of decades.

But, shop fabricated modules, which can be installed quickly in the field, are a very economical way to build equipment. So Focus Fuson technology, with factory made modules, may well be quite practical, even if the modules are then installed at larger facilities.

I would also expect them to be used at large industrial facilities, for both power and process heat. And again, in non-urban areas, and with constant supervison and site security.

[Brian H]

A pilot project using low grade heat and small amounts of electricity to desalinate water has been set up here in Vancouver, BC, and it seems to me to mesh perfectly with FF requirements.
http://www.economist.com/sciencetechnology/displayStory.cfm?story_id=14743791
and
http://www.saltworkstech.com/technology.php

In simple terms, it uses heat or sunlight to concentrate some seawater, then ionic gradients to run a kind of double loop of Cl and Na ions in opposite directions to seawater separately pumped to the process chamber. The resulting process requires 1 kwh per 1,000 liters of purified water (mainly for pumping, etc.) So a 5MW FF generator would be able to provide the energy to purify 24×5,000×1,000 120 million liters of water a day, or about 40 million USG, at a cost of around $300 plus any expense for (passive?) heat transfer/transport into the process.
Not tea bags!

[Rematog]

The other costs for the water purification system would be cost of capital for the funds used to build it and O and M costs. These O and M costs would include operating staff and maintenace parts and services. If you want an idea about staffing, maybe you could compare it to a water treatment plant.

I’ve seen several proposals from academic types for industrial plants. One that sticks in my mind, was for growing algae using the high CO2 fuel gas from the boilers.

They assumed O&M;to be a percentage of revenue. When you did the math, they would have a 1,000 acre algae farm operated by one person at night, 2 during the day, with a management staff of 2-3 people….Totally un-realistic (they didn’t take into account it takes a miniumum of 4 employee’s to provide 24-7 coverage of one person, and that if they work overtime to cover each others vacation, sickleave, training, holidays, etc, so you have to divide the operating staff by 4 to get actual people on site at any one time). Just based on safety, you don’t have someone alone on a site… second, too many jobs take more then one person… etc.

Then the management effort needed to meet H.R. requirements, environment requirements, safety requirements, planing, purchasing, etc. mean that 2-3 management people CAN NOT do the work needed.

Example, as a plant engineer for a small tile manufacturing plant (did this for three years during the utility down-turn in the 90’s), I attended a EPA seminar on toxic’s reporting. Plant’s of any significant size have to file detailed inventories, cradle to grave tracking, of all “toxics” (you can guess the EPA includes many, many things in this definition). The speaker, trying to bond with the paticipants, said that they understood how difficult compliance is, “especially for smaller sites, with only two or three environmental compliance staff”. The EPA feels that having 2 or 3 people on staff, full time, with degree’s, is insufficent to deal with the regulatory requirements. And they are not wrong.

I’m sure if I tried to do hard science research, I’d be way out of my depth… but someone with no industrial experience also is out of their depth when they handwave away the realities of maintaining and operating an industrial plant.

[Aeronaut]

Brian H wrote: A pilot project using low grade heat and small amounts of electricity to desalinate water has been set up here in Vancouver, BC, and it seems to me to mesh perfectly with FF requirements.
http://www.economist.com/sciencetechnology/displayStory.cfm?story_id=14743791
and
http://www.saltworkstech.com/technology.php

In simple terms, it uses heat or sunlight to concentrate some seawater, then ionic gradients to run a kind of double loop of Cl and Na ions in opposite directions to seawater separately pumped to the process chamber. The resulting process requires 1 kwh per 1,000 liters of purified water (mainly for pumping, etc.) So a 5MW FF generator would be able to provide the energy to purify 24×5,000×1,000 120 million liters of water a day, or about 40 million USG, at a cost of around $300 plus any expense for (passive?) heat transfer/transport into the process.
Not tea bags!

Cool idea except for their math. I agree with Rematog’s staffing numbers plus throw in the fact that this design would not scale well without using huge stainless steel pipes, pumps, and motors. So I’m reading a tradeoff between lots of jobs or lots of startup capital to make 40M g/day.

[Brian H]

Aeronaut wrote:

A pilot project using low grade heat and small amounts of electricity to desalinate water has been set up here in Vancouver, BC, and it seems to me to mesh perfectly with FF requirements.
http://www.economist.com/sciencetechnology/displayStory.cfm?story_id=14743791
and
http://www.saltworkstech.com/technology.php

In simple terms, it uses heat or sunlight to concentrate some seawater, then ionic gradients to run a kind of double loop of Cl and Na ions in opposite directions to seawater separately pumped to the process chamber. The resulting process requires 1 kwh per 1,000 liters of purified water (mainly for pumping, etc.) So a 5MW FF generator would be able to provide the energy to purify 24×5,000×1,000 120 million liters of water a day, or about 40 million USG, at a cost of around $300 plus any expense for (passive?) heat transfer/transport into the process.
Not tea bags!

Cool idea except for their math. I agree with Rematog’s staffing numbers plus throw in the fact that this design would not scale well without using huge stainless steel pipes, pumps, and motors. So I’m reading a tradeoff between lots of jobs or lots of startup capital to make 40M g/day.

Yeah, if you read carefully, I was referring to the cost of providing the heat and power from an FF generator, at the projected billing rates for such output. What it takes to run the purifier itself is another question altogether.

[Author]

Posts