Posts by ShneekeyTheLost

Quote from Gaxx

i don't think the Von Neumann machine works here, since we don't want a self replicating machine, we want a growing machine more like a living organism.
i still think it would be possible to realize it as a growing ring. you'd have to create start state where you use it as a normal framequarry, creating the "centre" of your ring. after that you split it into 8parts, moving into 4directions where 2parts are a side. those 2parts aren't connected, so you can move them apart and fill the space between them with new parts to let them grow. the middle between those 2parts must ever consist of frames, so they won't connect, but you need a reconfiguration mechanism to change that once a new part becomes the middle.

If you want a Von Neumann machine, use Computercraft. Crafty Turtles to build new turtles, mining turtles for everything else. Facility on RP frames to scrap and convert to UUM all the cobble and crap, and to store everything else. It would also need to be out in the open, and have a method of recharging the turtles from renewable resources, but theoretically possible.

The problem is that CRCS designs really aren't all that cost effective in terms of EU/t per resource investment until you run it for extensive periods of time, due to the number of cooling towers which effectively multiply your initial resource investment.

Quote from JstuffJr

I'm not so sure about that - I've been experimenting and tinkering a lot and it looks like I should be able to get some descent eu/t without using copper based components, so I'll only have to use copper for multicells and the chambers themselves, which drastically reduces the amount of copper.

OC Heat Vents require 16 copper each. I don't see how you are going to mitigate any significant amount of heat without them.

Quote from ShneekeyTheLost

My suggestion to you: Tunnel Bore covering level 50-35. That's where most of the copper will reside. Scrap the cobble.

Yeah I have multiple quarries running and a nether lava centrifuging station, but I hadn't thought about using a tunnel bore. Anyone have a link that explains making tunnel bores to a custom size? The only info I can find on wikis is on the railcraft tunnel bore which apparently only mines 3x3 :-/[/quote]
Use RP2 Redstone Frame Tubes connected to block breakers to make the face. Use a couple of frame motors in an inchworm drive. Set up a timer to run it. Hook up an EnderChest to the back of the face so all the input from the block breakers goes into it, and hook it up to your sorting system of choice. Hook up a Chunkloader to keep it loaded.

Quote from Requia

Look at the CRCS system thread then, with helium coolant cells you can avoid most of the copper cost.

Eh, multiple cooling towers probably blows his copper budget, though.

To the OP: Not gonna happen. Heat Vents require a bunch of copper. Nuclear material takes up a lot of heat. The more EU you produce, the more heat you have to mitigate, which means more copper.

My suggestion to you: Tunnel Bore covering level 50-35. That's where most of the copper will reside. Scrap the cobble.

Quote from Kenken244

Very nice, well written. A couple of your links are messed up or link to the wrong thing though, you might want to check that.

Ahh, thank you for letting me know. Everything should be pointing properly now. Let me know if anything remains screwed up. Those responsible have been re-assigned to reactor prototype testing.

So it is. Thank you, I felt horribly guilty about using so many NR's in that build.

You may wish to try the CRCS HVC reactor.

It consumes no copper, although it does have some initial startup costs. You will also want a single-chamber cooling tower also full of HVC's. Every 40 seconds, you will need to swap them. This can be automated with RP2 or CC.

At 440 EU/t, you will need eleven of them to hit your EU/t numbers.

Using Plutonium instead of Uranium will result in doubling your numbers, however it will require more cooling tower space and a much shorter micro-cycle time. In fact, you will need some four times the amount of cooling tower space, making it not economical for you.

There are several important numbers to keep track of when using a CRCS system which are not necessary in traditional Mk. I reactors. We'll go over these, why they are important, and get into some examples and how the numbers crunch.

One of the first numbers is a concept which nearly died out with the advance in nuclear engineering technology, and that is the Micro-Cycle time. Used in old-school Mk. III reactors, and revived again in CRCS systems, the Micro-Cycle represents how long the reactor can run before its components start melting. This number is easily found in any Reactor Planner software under the label 'Time Limit', found at the bottom of the same section the reactor components are kept. As an example, you can see on this DDoS reactor, the Micro-Cycle time is 179 seconds. You can also derive this number under the 'Timing' section, the very first number "Generation..." is your micro-cycle time.

The next important number is your cooling cycle time. This is derived from dividing your cooling capacity by your cooling tower's cooling per cell per second. Thus, a system using 60k Cooling Cells, with a 60 Cooling/cell/second tower, has a 1,000 cooling cycle time. Obviously, the higher your cooling per cell per second number is, the shorter your cooling cycle is, which is why the latests generation of cooling towers are so exciting. With a cooling of 96/cell/second, a 60k cooling cell has a cooling cycle time of a mere 625 seconds.

To derive the number of cooling towers you will need for a CRCS system, you will need to divide your cooling cycle time by your micro-cycle time, then multiply by the percentage of towers needed to hold the components. Therefore, in our examples, a generator tower with a micro-cycle time of 179 seconds has 8 cooling cells. However, the cooling tower can only handle 6 cells at a time. This means you will need 1.33 towers per cycle. Therefore, 625/179 = ~3.5 (rounding up for simplicity's sake) * 1.3 = 4.655. Now, you can't build half of a cooling tower, so you have to round that up to 5 cooling towers required to accomidate that reactor tower.

Now then, let's say the reactor tower got an upgrade with some OC vents and component heat exchangers, as shown here. This one has a micro-cycle time of 200 seconds. So crunching the same numbers, we end up a tiny fraction larger than 4. Now, since you have to have a brief pause to swap cooling cells, and give the OC vents time to cool, you can probably get away with 4 cooling towers, although you'd be cutting your safety margins down to slim pickings. So, despite the additional expense in the tower, you save far more by cutting down the number of towers needed by 1.

Now then, a major engineering challenge is to automate the exchange of the heat-laden components to the cooling tower(s) and cool ones back to the generator reactor. By far the most roubust system employs uses Redpower2 logic circuits.

A Timer is employed in conjunction with a State Cell to run the reactor. The Timer and State Cell are set to the same number, which should be a factor of your micro-cycle time. Thus, if you have a micro-cycle time of 200, these should be set to 20 seconds. If you have a remainder, that can be added to the state cell time. Thus, if you had a micro-cycle of 202, you set your timer to 20, and your state cell to 22. When the Timer pulses, it resets the State Cell's timer, then on the final cycle, the state cell finally gets to use those last two seconds.

Also connected to this timer is a counter. This will count the number of times the timer pulses. When it hits the desired number of pulses, it will activate, and stop the timer. It would be good practice to use an OR gate which accepts input from this timer, which is connected to a Toggle Latch that triggers the timer to be stopped.

Your thermal monitor and Nuclear Control System should also be able to send a signal through that same line, so if any heat buildup occurs in your reactor, it will automatically shut down the whole operation.

Next will be a Filter and a Retriever. If the system is transferring nine or fewer cells each way, then a signal from the Toggle Latch is all that is necessary to trigger the exchange. Simply have precisely the number of cells in each, and one pulse pulls the correct ammount.

**IMPORTANT**

Redpower2 machines differentiate between cells with a damage value and those without a damage value. Therefore, it is imparitive that you have components WITH a damage value in the Filter, to pull them out of the generator reactor, and components WITHOUT a damage value in the Retriever, to keep it from pulling still warm components back.

If you have more than nine cells in your generator reactor, you will need to employ a State Cell hooked up to a NOT gate and a Timer. When the signal from the Toggle Latch hits the State Cell, it will start a countdown. This will turn off the NOT gate, and allow the Timer to pulse. The duration of the state cell divided by the pulse rate of the timer will be the number of cells pulled in both directions.

You may need to employ cover strips to prevent unwanted connections in more compact designs.

Attached to your Retriever should be an in-line Item Detector connected to a counter. It will count the number of cells entering your reactor. Once it reaches the tally of the number of cells that are supposed to be in there, it will send a signal to the OR gate attached to the Toggle Latch. Assuming there isn't already a signal in the OR gate, say from your Nuclear Control system detecting heat or the cooldown period hasn't been reached yet, the OR gate will pulse, turning on the Toggle Latch and restarting the system. This will ensure that your reactor will not start without all cells physically present in your reactor.

As an alternative, you can use ComputerCraft to automate this system far easier. Use rp.setBundledOutput commands, have one color going to your reactor, and a seperate color going to both the retriever and the filter. You hook up your Nuclear Control killswitch to the computer. Set it up so that when it receives a redstone signal from that side, it automatically kills the reactor and initiates a cell cycle, then waits for the greater of either a full cycle time or the signal turning off before restarting.

While this is a much more compact and elegant design, it is not 'restart resistant'. When the server resets, so do the computers. Developing a program whch is restart resistant is theoretically possible, but I leave that up to the experts in the LUA field.

Thank you all for coming today. I hope you now have a deeper understanding of the strengths and weaknesses of a CRCS system, and what is involved in running and maintaining one. If you wish to keep abreast of the latest discoveries in the CRCS field, I suggest you keep an eye on the Nuclear Engineering newsfeeds.

Refreshments are available in Ballroom C next door. Representatives from HAYO Corp. and ShneekeyCraft Inc. are available to discuss what options are best for your situation, and have order forms on hand with special discounted offers for atendees of this seminar. You may also indicate on the guestbook if you wish to be kept abreast of new products from either of these companies.

Ladies and gentlemen, a moment of silence for the workers who gave their lives to bring us this new technology... okay, that oughta about do it. There will be a two hour recess until the next lecture. Thank you, and good day.

Hello, and thank you for attending the CRCS Seminar. As promised, electric thermometers and red alloy wiring kits are included in everyone's package as a part of the cost of the seminar.

I'm sure you are here to hear about the new CRCS systems, likely having heard such things as 'CASUC 2.0'. The truth is... far more complex. There are many engineering challenges involved with a CRCS system.

In fact, that does bring me to a point. The CRCS system is not one which is designed for beginners. There are many fault-points in any CRCS system. Running and maintaining a CRCS system will absolutely require an in-depth knowledge in the following fields: Nuclear engineering, logistical networking, pneumatic engineering, redstone wiring, logic circuitry, mechanical engineering, and possibly LUA coding as well.

If you don't have at least a solid understanding with reference materials for these topics, I would strongly advise you to take this final opportunity to depart with a partial refund now. A HAYO Corp. agent is just outside with design specifications for their Mk. I Reactor line, which are far easier to set up and employ, and will meet the needs of most consumers.

Anyone else? There's no shame in acknowledging your limitations.

All right, now let's get down to the brass tacks.

CRCS stands for Continuously Re-applied Coolant System, and its roots do stem from the old CASUC system, or Continuously Applied Single Use Coolant. However, as the only single-use coolant which works with current generation reactors consumes copious quantities of redstone or lapis, it is no longer economically viable for most situations. So alternatives were explored.

The basic concept of the CRCS is a radical departure from standard Nuclear Engineering protocols. Rather than try to dissipate the entirety of the heat generated from nuclear fission, components are used to store the heat. Before the components exceed safe levels, the reactor is temporarily disabled and the 'expended' components are transferred to one or more 'cooling towers' which have no fissionable materials but significant cooling potential.

This allows you to disperse your heat to a far larger heat sink, and enables the CRCS system to handle far higher heat levels than any traditional reactor to date, which means producing Energy Unit output far higher than any Mk. I reactor, capable of meeting or even exceeding old-school CASUC levels. The downside, however, is the massive infrastructure and initial investment of resources required to build such a system. As every cooling tower is effectively its own reactor, the costs involved are non-trivial. Having said that, most CRCS systems have a higher efficiency rating than a traditional Mk. I reactor, so maintenance costs are actually lower than one would expect. However, it would take a great deal of time to achieve the break-even point on initial cost versus maintenance cost savings.

The very first CRCS system was the old and antiquated DDoS Alpha reactor as seen here. It employed 60k Cooling Cells to store the heat, which were then transferred into the classic Coolmaster tower found on the following page.

While it was able to exceed even the EU output of a CASUC reactor, there were several problems with the system. First and foremost, it required 24 cooling cells per cycle, and with a cooling of 12-16 heat per cell per second, it required a huge number of cooling towers. Quite frankly, it was not economically viable.

Later DDoS reactors would scale it down to more manageable levels. Still exceeding the EU/t and Eff ratings of Mk. I reactors, but requiring far fewer cooling cells and far fewer cooling towers to accommodate them. Some designs also employed Neutron Reflector caps on the uranium rods to normalize heat dissipation values and increase efficiency, albeit with a higher maintenance cost, due to the neutron reflectors needing to be replaced periodically.

The next advance in CRCS technology came from an innovator by the name of Kenken244, who introduced the HVC (Heat Vent Cycle) reactor. He realized that while the Overclocked Heating Vent could only hold a fraction of the heat a 60k Cooling Cell could, they were self-cooling. This resulted in the original HVC design, found here.

Since OC Vents cool themselves 20 heat/second, employing them in a Coolmaster tower more than doubled, nearly tripled in fact, the cooling per component. The downside is the dangerously narrow safety margins. The designer suggested a one-second micro-cycle, and any sort of server lag could have disastrous consequences. However, the cost savings in having so many fewer cooling towers was quite attractive, making it a very viable option.

You can also use an empty reactor as a cooling tower for an HVC unit, which dramatically slashes costs by doubling the space available per cooling tower, as well as the cost savings in all the component heat vents you don't need. While it cools down slower, it is also able to be far more compact.

HAYO Corp. developed the next innovation in CRCS systems, the hybrid generator reactor. In essence, they employ cooling units in the generator reactor in addition to 60k coolant cells to increase the micro-cycle time. Focusing on reducing the copper expenditures required on the quad-cell reactors, they produced the CRCS Mk-1-EB-3.93 and the CRCS Mk-1-EA-4.5 reactors.

The significant advantage of the first is that it only uses regular Uranium Cells, instead of dual or quad cells, meaning far less copper and tin is consumed per cycle, further reducing maintenance costs. In fact, if one has access to certain bee species which produce tin and uranium, the entire system is completely renewable. The second system produces more EU/t, and still doesn't rely on quad cells which the DDoS reactors employ, resulting in a noticable savings in copper.

In rebuttal, ShneekeyCraft Inc. released an HVC design here which used a single Coolmaster Tower. With a 50 second micro-cycle, HVC may have found a new life in budged CRCS reactors.

The other truly revolutionary advance which HAYO Corp. published was a radically new cooling tower design philosophy. Kenken244 came up with the original design, which uses a fair amount of gold in the cooling tower in OC vents and component heat exchangers, as shown here. However, considering it gets a revolutionary 95 cooling per cell per second, it is well worth the expense. The only problem is the limited capacity. Most of these ultra-cooling towers only have a capacity for 4-6 cells at a time, requiring multiple cooling towers per cycle for some of the larger reactors.

These innovations, in turn, caused a radical re-design of the DDoS reactors, which also started employing OC vents and component heat exchangers to push the micro-cycle time.

Quote from skavier470

you will need 7 of the new C00ld0wn® REV.2e-6///568/95
to cool DDoS 12QUv2.0
and 3.4 C00ld0wn® REV.2e-6///568/95
for this DDoS 6QUv2.0 Pocket Edition.

Ahh, that was the problem! Oh well, at least the crater exposed a vein of Copper.

Currently, I'm working on improved heat dissipation within the generator-reactor itself. Unfortunately, that seems to be limited to the component heat vent's maximum capacity, which is rather disappointing. I've gotten my micro-cycle time to 200 seconds, but I fear more inventive re-engineering is going to be necessary to push that any further.

I have some ideas, but they are still in the designing stages.

DDoS-2.1 4QU/357 (Pocket Reactor)

DDoS-2.1 4UQ-NR/228

DDoS-2.1 6QU-NR/200

DDoS-2.1 6QU-NR/648

Quote from Shakie666

Don't mean to spam, but I asked on the previous page what such a setup would look like (with the counter attached to the item detector + everything). I repeat it here because i've tried a few setups but none of them seem to work.

Ahh, right.

The easiest way is via ComputerCraft. The rp.setBundledOutput comand can be used to turn on and off different channels.

Then you run bundled cabling, with one color going to the generator-reactor. One color runs to the generator-reactor. The other color goes to the retriever and filter.

With the latest generation reactor, which has more than nine cooling cells, you have one cell that has no damage bar in the retriever and one that has a damage bar. Then you pulse that color a number of times equal to the number of cells in the reactor. Then a brief pause to ensure that all of the cells are transferred and none of the components have any residual heat, then the cycle starts over again.

The downside to this is a) knowledge of LUA (which really isn't that difficult), and b) when the game resets, so do the computers. There is a way to make it reset-resistant, involving the startup command.

The other way is to use RP2 logic circuits. It's immune to resets, but isn't as compact.

Either way, this is not a build intended for new users to play with, because there are many potential fault-points. I would strongly suggest a typical Mk I reactor until you are more familiar with RP2 logic gates and machines.

Quote from Shakie666

Wouldn't you still be better off with the 360k cooling cells though, since you'd have longer micro-cycles, meaning less down time for the reactor, meaning higher effective eu/t?

Yes. Because the vents and heat exchangers are heating up, you'll have to have a cool down period every micro-cycle, so the fewer micro-cycles, the better it is. However, you will need an extra two cooling towers, which are kinda pricey, so it's up to the individual to determine if it is cost effective for them.

NEW AND IMPROVED!

With the latest in advances in cooling tower technology, ShneekeyCraft Inc. has introduced a new line of reactors which takes full advantage of the C00ld0wn® REV.2d-6///628/104.3 unit.

First, we have our DDoS 6QUv2.0 Pocket Edition.
EU/t: 720
Eff: 6
Micro-cycle time: 200 seconds
Total EU output per full cycle: 144 Million
Number of C00ld0wn® REV.2d-6///628/104.3 units required: 3

While it produces less EU/t than the typical, it takes full advantage of the 575 cooling cycle time of the REV.2d. It is advised to pause between micro-cycles to give the other components time to cool down before restarting. And it only needs three cooling towers.

However, if that isn't enough for your needs, we have something a little... larger.

DDoS 12QUv2.0
EU/t: 1440
Eff: 6
Micro-cycle time: 197 Seconds
Total EU output per cycle: 288 Million
Number of C00ld0wn® REV.2d-6///628/104.3 units required: 6

As you can see, it cuts the safety margin a little closer to the bone, but you've still got a five second leeway. Double your EU output, double your number of C00ld0wn towers you need. However, the energy output is certainly impressive.

Now, for those with access to GregTech components, there are some interesting options available.

Replacing the Quad Uranium with Quad Plutonium Cells outputs 2880 EU/t for a total of 1,152,000,000 EU per cycle. Yes, that's right, over one BILLION EU per cycle! Of course, with an 84 second micro-cycle time, you'll be needing 14 REV.2d reactors to keep up with the heat. Replacing with 360k cooling cells unfortunately hits the melting point of the heat exchangers, thus actually ends up in needing two more cooling towers.

Using quad thorium cells has an interesting reaction. On the one hand, the micro-cycle time exceeds the cooldown time, meaning you only need two towers to handle all 12 cooling cells. However, at 288 EU/t, there are probably more cost-effective methods of generating this energy level output. Still, 288 Million EU out of a total cycle isn't a bad deal.

The pocket plant is scaled as you might expect. You'll need 7 REV.2d cooling towers for the Plutonium variant, which produces 1440 EU/t.

Ironically, each of the C00ld0wn Rev.2 reactors with a cooling of 100/cell can also double as a Mk. I reactor by replacing the cooling cells with quad-uranium cells. Granted, they are not as efficient as, say, the Beast... but it will at least work.

Hmm... There will shortly be a new generation of DDoS generator to take advantage of the new system's capacity and efficiency. This could get quite... interesting.

Quote from skavier470

#### ####
<1931y34>

We have made some vast improvements on our C00ld0wn® system.
Kenken244, one of our external Developers, helped with some designs which are now at our HAYO lab 17.
For now we can't publish the new designs as they are not yet finalized.
We are just letting you get a glimpse on the HAYOish naming system we gave them:
C00ld0wn® REV.2b-8///544/68
C00ld0wn® REV.2c-6///600/100
C00ld0wn® REV.2c-4///400/100
C00ld0wn® REV.2c-2///200/100
C00ld0wn® REV.2d-6///628/104.3

some of the more sophisticated engineer will get what we have done.....

Ps: just go for it and guess for what the numbers stand, I will post the Designs when you got it right.....

Display More

Connecting to external CPU sources...
Activating ShneekeyCraft .ORGv2.1... 100%
Connection established 0% loss (0 packets of 256 packets)
Activating secondary systems... 2,317 sources responding




HELLO, DAVE
:~$ Stop calling me Dave
AS YOU WISH, DAVE
:~$ sudo wget "http://www.HAYO.corp.com/News/20022013.htm" "HAYO.htm"
[sudo] password for [REDACTED]:
[SECURITY PROTOCOL #69 DATA EXPUNGED FOR [REDACTED] LINES]
download complete




:~$ /.decrypt HAYO.htm
Running...
... 12%
... 16%
... 24%
... 32%
... 49%
... 56%
... 62%
... 77%
... 83%
... 96%
... 100%
... Cross-checking algorithm 100% SUCCESS
:~$ print decrypted_HAYO.htm

FORMAT: (a)(b)-(c)///(d)/(e)

EXAMPLE FROM INITIAL DATA SAMPLE: C00ld0wn® REV.2b-8///544/68

(a) = [C00ld0wn® REV.2]
(b) =[ b ]
Partition code: -
(c) = [8]
Partition code: ///
(d) = [544]
Partition code: /
(e) = [68]

(a) = Cooling Tower designation. Input type = expended cooling cells. Probability 98% +- 2%
(b) = generation type. Possible indicator of material type or component type. Probability 53% +- 10%
(c) = Cooling Cell capacity. Probability 89% +- 6%
(d) = Total cooling per second per tower. Probability 87% +- 9%
(e) = Cooling per second per cell. Derived value (d)/(c). Probability 95% +-5%

DECRYPTION COMPLETE

Quote from Requia

Rick has stated, twice, that he will no longer be maintaining the reactor design sticky, and that somebody else needs to take over.

There' has been zero discussion on this in the thread, I don't believe many people saw his posts. I am willing to take over, but I will understand if others feel that someone else should, in which case we need a different volunteer.

When was the last time the list *needed* to be updated? I mean, sure... there might be one or two people with some novel ideas, but unfortunately no one has really had anything to meaningfully contribute to that thread in quite some time. Generally, it's yet another newbie who is absolutely convinced that he has the best reactor design EVAR... without consulting the original post which generally has one or more reactors which is strictly superior.

Quote from Shakie666

I would just like to mention I tried building one of these myself, following the guide in the video on page 2. It worked perfectly, except when it came to putting the cells back in the reactor. Using the redstone setup in the video, the reactor came back on before all the fresh cells had been replaced in the reactor (i.e. they were still travelling through the pipes when the reactor came on). How can I make it so there is a longer delay before the reactor turns back on?

Well, there's actually several ways to go about doing this, depending on how simple or complex you wish to go.

If you *really* want to be safe, you can use an in-line Item Detector attached to the Retriever pulling items into the reactor. This, in turn, is hooked up to a counter. As items go through the Item Detector, it pulses the Counter. Once all of the cooling cells have passed through, the counter lights up the AND gate to turn the reactor back on. That way, no matter how slow they take, it won't turn on until all the cooling cells are in the reactor.

Also, that reactor design is an outdated version. A more up-to-date version is found here. An updated cooling tower is found here with a 60 cooling per coolant cell (five times better than the old version). This means you will need a total of six such cooing towers, at a 175 second micro-cycle, to keep up with the heat.

Mmm... I see your point.

I'll see if I can tinker with the DDoS vDU a bit and get it to a 1,000 second micro-cycle. Then I will be able to use a single C00ld0wn REV.2b cooling tower, which will make the costs much more viable. So far, I have been unable to push it beyond 720 seconds, however I shall continue experimenting and see what I can come up with. Besides, I think I've made the creepers abandon the area out of 'crater envy'. This interesting side-effect bears more research as well...

I did find a rather novel design shown here. Using a 500 second micro-cycle, you would only need a single REV.2b cooling tower. However, there would be approximately a three minute cooldown period between cycles to allow the components to cool off. So it would effectively function as a Mk. III reactor.

The thick neutron reflectors are designed primarily to ensure even heat distribution over all the cooling cells. Otherwise the central cooling cells end up heating up much faster. The increase in efficiency, EU/t, and overall EU is merely fortuitous. However, you do have a valid point about needing to replace them periodically, and that not everyone wishes to consume resources.

Ahh, excellent. You have certainly made certain design improvements on my DDoS/Cooling Tower concept. However, there's still a few toys from ShneekeyCraft Inc. which you might wish to consider.

Now, I can understand your CRCS plant having built in partial cooling to compact the design and having fewer cooling towers required, however I can beat your efficiency AND your EU/t at no additional increase in cooling cycle, and do it in a single chamber reactor.

DDoS Reactor Type DU

Now granted, I use more coolant cells than you do, which means more cooling towers.

At 60 cooling per second per cell, you are looking at at precisely 1,000 seconds for maximum cooldown. Now, if you pull it back a second, to give yourself a margin of error, you'll have a 500 second micro-cycle. This means you'll need 4 of your REV.2a or 2 of your REV.2b cooling towers.

You may have accidentally leaked a prototype with your first link, though, because it has LHZ's in it. However, this can be addressed with an even more simple solution, the HVC.

Now, the reactor itself looks like this with some impressive stats.

Efficiency: 4.0
Heat Generated: 640
Internal Cooling: 320
EU/t: 320
EU Total: 64,000,000

The advantage of this system is the 'cooling tower' is simply an empty single-chamber reactor. And you'll only need one of them. You can set it up to a very convenient 30 second micro-cycle time, which comes with a very generous 20 second buffer, and just swap them out every so often.

No, you are not going to be able to remove EU loss from your cables.

Have a nice day