Latest / Mechanical Engineering Made Simple / (#31) The Engineering Battle Against Air and Water Pollution

Transcript

(#31) The Engineering Battle Against Air and Water Pollution · Mechanical Engineering Made Simple

• 0:00
  Welcome back, Design Junk Houston Mechanical Engineering
• 0:02
  Made Simple. Today we're taking a deep dive
• 0:04
  into the engineering solutions developed really since the
• 0:08
  Industrial revolution. We're tackling a huge problem.
• 0:11
  How do we manage and control stationary source pollution?
• 0:15
  You know, the stuff hitting our air and water.
• 0:18
  Yeah, and it's not just history, right?
• 0:19
  It's a constant engineering challenge.
• 0:21
  Today we're looking at big industrial activities.
• 0:24
  Think power plants, metal processing, manufacturing.
• 0:26
  These are major stationary sources pumping out pollutants
• 0:30
  continuously. The scale is just.
• 0:33
  Massive, and the engineering challenge that comes with it.
• 0:35
  Exactly. Colossal What's?
• 0:36
  Really interesting is how much the law drives innovation.
• 0:39
  Here you've got the EPA, the Clean Water Act, RCRA.
• 0:43
  These aren't suggestions. Oh, definitely.
• 0:45
  Not They set hard limits. They force engineers to get
• 0:47
  creative, turning simple disposal into really complex
• 0:50
  chemistry and flow control systems.
• 0:52
  So the regulations set the bar and engineers figure out how to
• 0:56
  clear it efficiently and hopefully cost effectively.
• 0:58
  That's the name of the game. Performance first, then optimize
• 1:01
  the cost. OK, let's kick things off with
• 1:03
  air pollution. The big challenge there is
• 1:06
  managing the six criteria pollutants.
• 1:08
  These are the ones defined by health standards, the culprits
• 1:11
  behind smog and acid rain. We're talking carbon monoxide,
• 1:14
  Co, nitrogen dioxide, No2, ozone, particulate matter
• 1:20
  specifically PM 10 so less than 10 micrometers, sulfur dioxide,
• 1:23
  SO2, and lead PB. Right.
• 1:26
  And for stationary sources, controlling these usually falls
• 1:29
  into three main buckets. 3 engineering philosophies if you
• 1:32
  will. First up is pre process control.
• 1:35
  Basically clean the fuel before you burn it, then cool cleaning
• 1:38
  removing sulfur or ash beforehand.
• 1:40
  Makes sense. Get rid of the problem before it
• 1:42
  starts. Then you have process
• 1:43
  modification. This is about changing the
• 1:45
  actual combustion itself, or maybe the manufacturing process
• 1:48
  to stop the pollutant from even forming.
• 1:50
  Like the Lonox burners we hear about.
• 1:52
  Exactly like Lonox burners. That's a prime example.
• 1:55
  Change the flame to prevent novaks.
• 1:57
  And the third way, that's the classic post process control,
• 2:00
  sometimes called end of pipe. Here the pollutant is already
• 2:03
  formed and you clean it up from the exhaust gas after combustion
• 2:06
  flue gas desulfurization systems.
• 2:08
  The scrubbers fit right in here. Got it.
• 2:11
  Let's dive into the first big one then.
• 2:12
  Sulfur dioxide, SO2. This is mostly a fuel borne
• 2:16
  pollutant, you said. That's right, the sulfur is
• 2:18
  chemically part of the fuel, like coal or heavy oil, before
• 2:22
  it even gets near the furnace. So the preprocess option seems
• 2:25
  obvious. Switch fuels maybe?
• 2:27
  Yep, fuel switching is one way. Go from high sulfur coal like
• 2:31
  you find in the eastern US to lower sulfur western coal.
• 2:34
  Or try cleaning the coal you have.
• 2:37
  But there's a catch, a big one. Even with the best coal cleaning
• 2:40
  tech for those common high sulfur eastern coals, you
• 2:43
  typically can't get more than about 30% of the sulfur out.
• 2:46
  Not without spending a fortune, anyway.
• 2:48
  Wow, only 30%? So like 70% or more is still
• 2:51
  going into the furnace. Pretty much, which means the
• 2:53
  main fight against SO2 often moves downstream to post process
• 2:58
  control. We're talking flue gas,
• 3:01
  desulfurization, FG, DS scrubbers.
• 3:04
  Right scrubbers. And there's a choice there
• 3:06
  between wet and dry systems. Correct, wet scrubbers are
• 3:09
  generally the high efficiency champs.
• 3:11
  They can consistently pull out 90% or more of the SO2.
• 3:15
  They usually use a slurry of lime.
• 3:17
  That's Cao or limestone case AO3.
• 3:20
  This slurry reacts chemically with the SO2 gas.
• 3:22
  To form what exactly? It forms calcium sulfite case O3
• 3:26
  or calcium sulfate case O4 gypsum, basically if it fully
• 3:29
  oxidizes. But I remember hearing those
• 3:31
  reaction products can be tricky, they cause problems.
• 3:34
  Right. Oh, they absolutely can.
• 3:35
  Yeah. Calcium sulfite and sulfate
• 3:37
  aren't very soluble in water. So the big operational challenge
• 3:40
  is managing this slurry that forms and, crucially, preventing
• 3:43
  scale deposition. Scale like limescale in a
• 3:46
  kettle. Sort of, but on an industrial
• 3:48
  scale. It's this sticky, solid material
• 3:50
  that builds up inside the pipes, their reactors.
• 3:52
  It can choke the whole system if you don't manage it constantly.
• 3:55
  It's a maintenance nightmare. OK, so that's wet scrubbers.
• 3:58
  What's the deal with the alternative, the dry scrubbers
• 4:01
  or spray dryer absorbers? Dry scrubbers or SDA's.
• 4:06
  They trade some of that top end efficiency for lower costs, both
• 4:10
  capital to build them and operating costs.
• 4:12
  Plus they produce a dry powder waste which is much easier to
• 4:16
  handle than a wet slurry. And the efficiency?
• 4:18
  Typically lower, yeah, maybe in the range of 70 to 90%, SO 2
• 4:22
  removal. OK, now you mentioned a
• 4:24
  counterintuitive insight about cost.
• 4:27
  Yes, this really gets to the heart of engineering trade-offs.
• 4:31
  When you look beyond the sticker price, the installation cost,
• 4:35
  and instead measure the true performance metric, the cost per
• 4:38
  ton of SO2 removed, something interesting happens.
• 4:41
  What's a The cost often becomes surprisingly comparable across
• 4:44
  different technologies, even between the high efficiency wet
• 4:47
  scrubbers and some lower efficiency options.
• 4:49
  Wait, hold on. If the cost per ton removed is
• 4:51
  similar, why would anyone choose a technology like say furnace
• 4:55
  sorbent injection which might only give you 50 or 70%
• 4:58
  reduction? Why not always go for the 90
• 5:00
  plus percent? That's the perfect question.
• 5:02
  It comes down to the total amount of reduction you actually
• 5:05
  need to meet the regulations and also capital risk.
• 5:08
  Those cheaper, maybe lower efficiency systems?
• 5:11
  They're often easier and less risky to retrofit onto an
• 5:14
  existing plant. They're suitable for smaller
• 5:16
  facilities or maybe situations where you only need, say, a 60%
• 5:20
  cut to comply. I see.
• 5:23
  So if the law only requires A moderate reduction, you might
• 5:25
  pick the lower capital cost option even if the per ton cost
• 5:28
  is similar. Exactly, but if you're near a
• 5:31
  sensitive area, maybe a National Park, and you need 95% or even
• 5:35
  higher removal, then you really have no choice.
• 5:37
  You have to bite the bullet and pay the higher front cost for
• 5:40
  the high efficiency wet scrubber.
• 5:41
  The total pounds or tons of pollution removed dictates the
• 5:45
  choice driven by the regulations.
• 5:46
  OK, that totally reframes how you'd approach that decision.
• 5:49
  Let's shift gears now to the second big air challenge.
• 5:52
  Oxides of nitrogen, No VIX. That's NOx, usually a mix of NO
• 5:56
  and No2. Unlike SO2, which is mostly in
• 5:59
  the fuel, no X is processed derived right created by the
• 6:02
  combustion itself. Precisely.
• 6:04
  It's formed during the burning process, not just carrying with
• 6:07
  the fuel. All the fuel nitrogen does
• 6:09
  contribute. This means controlling it
• 6:11
  requires messing with the flame itself.
• 6:13
  We usually breakdown not cost formation into 3 pathways.
• 6:17
  The big one, especially in high temperature boilers, is thermal
• 6:21
  knocks. Thermal meaning heat related.
• 6:24
  Exactly, it's super dependent on temperature.
• 6:26
  When you get really high flame temperatures, the nitrogen N2
• 6:30
  and oxygen O2, they're just naturally present in the
• 6:32
  combustion air react with each other.
• 6:35
  The hotter it gets, the more thermal knocks you make.
• 6:37
  It increases exponentially. Wait, what's the second pathway?
• 6:40
  That's fuel knocks. This comes from nitrogen
• 6:42
  compounds that are chemically bound within the fuel itself.
• 6:46
  Coal is a big source of this. In many uncontrolled coal
• 6:49
  flames, this is actually the dominant source of knockoffs.
• 6:52
  Got it. And the third.
• 6:53
  The third is called Prompt Knocks.
• 6:55
  It forms very quickly right near the flame front, involving
• 6:57
  hydrocarbon radicals in large utility boilers.
• 7:01
  It's usually a pretty small slice of the total Knox pie
• 7:03
  compared to thermal and fuel knocks.
• 7:05
  So for control, the key take away seems to be manage the
• 7:09
  temperature and manage the oxygen.
• 7:11
  Hit those two and you reduce the two biggest knots.
• 7:15
  Sources. You nail it.
• 7:17
  Lower the peak flame temperature.
• 7:18
  Limit the oxygen available right where the fuel is burning
• 7:21
  fiercely and you can significantly cut down Knox
• 7:24
  formation. Which brings us back nicely to
• 7:27
  process modification and those Lonox burners L and BS you
• 7:31
  mentioned earlier, they're the go to prevention method.
• 7:34
  Absolutely. L and BS are really clever bits
• 7:36
  of fluid dynamics and combustion engineering.
• 7:39
  They work by carefully controlling how the fuel and air
• 7:42
  mix. It's called stage combustion.
• 7:44
  By slowing down that mixing, they create a bigger, less
• 7:46
  intense flame which lowers the peak temperature that directly
• 7:50
  tackles the thermal NOx formation.
• 7:52
  And the fuel NOx? They address that too
• 7:54
  critically. LM BS create zones within the
• 7:56
  flame where there's less oxygen available right when the
• 7:58
  volatile parts of the fuel are first burning off.
• 8:01
  Less oxygen there means less fuel NOx gets formed.
• 8:03
  It's a really elegant preventative solution.
• 8:05
  Stop it before it starts. Very neat, but what if
• 8:08
  prevention with LM BS isn't enough?
• 8:10
  Or maybe you're trying to upgrade an older power plant
• 8:12
  that wasn't built with? Them right Then you have to look
• 8:14
  at post combustion controls. These are methods that destroy
• 8:17
  the nuox after it's already formed in the flue gas, usually
• 8:21
  by reacting it with something. Abusing chemical reagents.
• 8:23
  Exactly. The two main technologies here
• 8:25
  are selective non catalytic reduction SNCR and selective
• 8:30
  catalytic reduction SCRSNCR involves injecting ammonia
• 8:33
  that's NH3 or sometimes urea directly into the hot flue gas.
• 8:38
  There's a specific temperature window.
• 8:39
  It needs quite high, typically 870 to 11150°C for the reaction
• 8:45
  to work well. And how effective is SNCR?
• 8:48
  It's decent you can get maybe it's a 75% in a wax removal.
• 8:52
  The big advantage is lower initial capital cost compared to
• 8:55
  SCR. But the downside is.
• 8:57
  High operating cost. You're constantly buying and
• 8:59
  injecting ammonia or urea. Those region cost added
• 9:02
  significantly over the life of the plant.
• 9:04
  OK, so that's SNCR. What about SCR selective
• 9:06
  catalytic reduction? SCR is the high performance
• 9:09
  option. It can often achieve over 90%
• 9:11
  nonox reduction, sometimes even higher.
• 9:14
  It uses the same basic chemistry reacting nonox with ammonia, but
• 9:18
  it employs A catalyst usually based on titanium or vanadium
• 9:22
  oxides. And the catalyst lets it work at
• 9:24
  lower temperatures. Exactly because of the catalyst,
• 9:26
  the reaction happens efficiently at much more manageable
• 9:28
  temperatures, typically around 315 to 370°C.
• 9:33
  Sounds great, but I'm guessing there's a cost catch again.
• 9:35
  Oh yeah, the catalyst itself is the major investment hurdle.
• 9:38
  It's expensive stuff. Initial cost can be up to what,
• 9:41
  $14,100 per cubic meter of catalyst and the catalysts do
• 9:45
  grades overtime, so you have periodic replacement costs too.
• 9:48
  Yeah, it's a significant capital outlay.
• 9:50
  So a clear tradeoff. SNCR is cheaper upfront but
• 9:54
  costs more to run. SCR costs a lot initially but
• 9:57
  gets higher removal and might be cheaper long term depending on
• 10:01
  Regent prices and removal needs. That's a good summary.
• 10:03
  The choice depends on the specific plant, the regulations
• 10:06
  it has to meet and the overall economics.
• 10:09
  OK, fascinating stuff on air pollution control.
• 10:11
  Let's pivot now, smoothly, hopefully over to the liquid
• 10:16
  world water pollution control. This field has also changed
• 10:19
  dramatically, hasn't it, moving beyond just making rivers look
• 10:22
  less gross. Yeah, absolutely.
• 10:24
  It shifted focus towards really managing public health risks and
• 10:27
  ecological impacts, again driven by regulations like the Clean
• 10:31
  Water Act and RCRA for hazardous waste.
• 10:34
  And when we talk wastewater treatment, it's important to
• 10:36
  distinguish between municipal, you know, sewage from homes and
• 10:40
  businesses and industrial wastewater.
• 10:42
  Their goals can be quite different.
• 10:44
  How so? What's the main goal for
• 10:46
  municipal treatment? For municipal plants, the
• 10:48
  primary job is reducing BOD. That's by chemical oxygen
• 10:51
  demand, basically the food for bacteria in assess suspended
• 10:54
  solids. The legal minimum requirement in
• 10:57
  the US is typically an 85% reduction of both BOD and S S
• 11:01
  coming into the plant, or sometimes it's stated as an
• 11:04
  effluent limit, like getting the treated water down to 30
• 11:07
  milligrams per liter or less for both.
• 11:09
  And how do they achieve that? Mostly biological processes.
• 11:12
  Yes, heavily reliant on biology. The workhorse is the activated
• 11:17
  sludge process, or maybe trickling filters.
• 11:20
  You're essentially harnessing natural microorganisms, giving
• 11:23
  them lots of oxygen and letting them eat the organic pollutants
• 11:26
  in the wastewater. Is natural enough, but there's a
• 11:28
  byproduct, isn't there? Sludge.
• 11:31
  Sludge. Yes, the bane of many a
• 11:33
  wastewater engineers existence. Well, the volume of sludge
• 11:36
  produced is relatively small compared to the huge volume of
• 11:39
  water treated. Managing it is incredibly
• 11:42
  complex and expensive. What makes it so difficult?
• 11:45
  Well, first it's mostly water, so you have to thicken it, then
• 11:48
  mechanically dewater it using centrifuges or belt presses to
• 11:51
  get the solids concentration up, maybe to around 30% solids.
• 11:55
  But then you have to stabilize it.
• 11:57
  This usually involves digestion, either anaerobic without oxygen
• 12:00
  or aerobic with oxygen. Why stabilize it?
• 12:02
  What does that do? Stabilization reduces pathogens,
• 12:05
  odors, and the tendency to decompose further.
• 12:08
  Remember, this sludge can contain disease causing
• 12:11
  organisms and sometimes accumulated heavy metals or
• 12:13
  other nasties. You have to process it to meet
• 12:16
  strict regulations before it can be safely disposed of in a
• 12:19
  landfill or ideally beneficially reused as biosolids.
• 12:23
  Basically a regulated fertilizer for agriculture.
• 12:26
  It's a whole process chain. Focus on volume reduction and
• 12:29
  making it safe. A constant battle, it sounds
• 12:31
  like. OK.
• 12:31
  Let's move beyond municipal treatment to the really tough
• 12:34
  stuff, advanced industrial wastewater and hazardous waste
• 12:38
  streams. Here, the pollutants aren't just
• 12:40
  simple organics. They can be complex chemicals or
• 12:42
  toxic metals. Right.
• 12:44
  For these you often need more specialized, sometimes physical
• 12:46
  or chemical separation methods. Standard biological treatment
• 12:50
  might not cut it. What are some key technologies
• 12:53
  here? Well, one widely used technique,
• 12:54
  especially for removing dissolved organic compounds, is
• 12:57
  activated carbon absorption, usually using granular activated
• 13:01
  carbon, or GAC. The principle is pretty simple.
• 13:05
  Activated carbon is incredibly porous.
• 13:07
  It has a massive internal surface area, sometimes hundreds
• 13:10
  or thousands of square meters per gram.
• 13:12
  Contaminants essentially stick to this huge surface area
• 13:15
  through physical absorption. It's great for polishing water,
• 13:17
  removing trace organics that cause taste, odor or toxicity.
• 13:21
  OK, the source is mentioned an equation for this.
• 13:23
  The frame link isotherm one Q is K1C.
• 13:26
  Can you break that down conceptually?
• 13:28
  What does Q = K sub 1 * C to the power of 1 / n tell an engineer?
• 13:33
  Sure. So dollars there represents the
• 13:35
  mass of the contaminant absorbed onto the carbon per unit mass of
• 13:38
  carbon. Think of it as the carbon's
• 13:40
  capacity. $2.00 is the concentration of the contaminant
• 13:44
  remaining in the water. K1 in dollars are constants that
• 13:48
  depend on the specific carbon and contaminant.
• 13:51
  Conceptually with that equation usually shows is diminishing
• 13:54
  returns. As the concentration the water
• 13:56
  gets lower and lower, the amount collars a carbon can still pick
• 14:00
  up decreases significantly. He gets harder and harder and
• 14:03
  takes more and more carbon to remove those last traces of the
• 14:06
  pollutant. So you hit a point where it's
• 14:08
  just not efficient to keep loading that same carbon bed.
• 14:12
  Exactly, which is why engineers often use multiple GA columns in
• 14:16
  series. You load up the first column
• 14:18
  until the contaminant starts to breakthrough the end.
• 14:21
  Then you take that first column offline for regeneration or
• 14:23
  replacement, move the second call into the lead position and
• 14:26
  add a fresh column at the end. Maximizing the use of the carbon
• 14:29
  before breakthrough makes sense. You squeeze every bit of
• 14:32
  capacity out of that expensive carbon.
• 14:34
  Another really important technique, especially for
• 14:36
  metals, is ion exchange. Ion exchange like in a home
• 14:39
  water softener. Similar principle, but often
• 14:42
  more sophisticated for industrial use.
• 14:44
  It's great for recovering valuable metals from waste
• 14:46
  streams or removing toxic ionic contaminants like heavy metals,
• 14:50
  cadmium, lead, chromium, etcetera.
• 14:53
  How does the basic chemistry work there?
• 14:55
  It relies on special resin beads that have charged functional
• 14:57
  groups. For removing positive metal ions
• 15:00
  conjugations, you'd use acacian exchange resin.
• 15:04
  A common form might start with hydrogen ions attached.
• 15:07
  Let's call the resin or dollar tech stage.
• 15:09
  Now when the wastewater containing a metal ion, say tech
• 15:12
  that's M plus dollar, flows through the resin, grabs the
• 15:15
  metal ion, and releases a hydrogen ion, the reaction is
• 15:18
  basically Texas test plus Cus plus TX plus plus test plus plus
• 15:22
  plus plus X plus dollar. OK, so R sub one H + M plus
• 15:25
  tends to R sub one M + H plus is the resin swaps the hydrogen for
• 15:29
  the metal. Precisely.
• 15:30
  The metal stays on the resin and the water leaves cleaner.
• 15:33
  But eventually the resin gets full of metal ions, right?
• 15:35
  It does. That's the exchange capacity.
• 15:37
  Once it's exhausted, you have to regenerate it.
• 15:39
  You typically do this by washing the resin with a strong acid
• 15:42
  solution. The high concentration of
• 15:45
  hydrogen ions in the acid forces the reaction to go backward,
• 15:48
  kicking the metal ions off the resin and putting the hydrogen
• 15:51
  ions back on to get your R dollar resin back ready for
• 15:56
  another cycle, and you end up with a concentrated stream of
• 15:58
  the recovered metal. A very selective way to pull out
• 16:01
  specific ions. Neat.
• 16:03
  OK, one more key area. Membrane technologies.
• 16:06
  Reverse osmosis. RO seems like a big one.
• 16:09
  RO is incredibly powerful. It uses a semi permeable
• 16:12
  membrane that allows water molecule tools to pass through
• 16:15
  but rejects almost everything else.
• 16:16
  Dissolved salts, organics, even tiny inorganic ions.
• 16:20
  It's used for desalina tion, creating ultra pure water and
• 16:23
  concentrating waste streams. How does it work against normal
• 16:26
  osmosis? Well, normal osmosis wants water
• 16:29
  to flow from a dilute solution to a concentrated 1 to equalize
• 16:32
  things. Reverse osmosis forces the water
• 16:35
  the other way, from the concentrated side, like salty
• 16:37
  water, to the dilute side, pure water, by applying pressure.
• 16:41
  And the pressure needed has to be higher than the natural
• 16:44
  osmotic pressure of that concentrated stream you're
• 16:47
  creating. Yeah.
• 16:47
  We're talking significant pressures requiring powerful
• 16:50
  pumps and robust membrane housings.
• 16:53
  Sounds energy intensive. What are the big operational
• 16:57
  headaches with RO? 2 main enemies fouling and
• 17:00
  scaling Fouling is the general term for stuff physically
• 17:03
  clogging the membrane surface suspended solids, biological
• 17:06
  growth, organic molecules sticking to it and.
• 17:08
  Scaling. Scaling is more specific.
• 17:10
  It's when dissolved minerals in the feed water become too
• 17:12
  concentrated as water passes through the membrane and they
• 17:15
  precipitate out as hard scaled directly on the membrane
• 17:18
  surface. Calcium carbonate KCO 3 is a
• 17:21
  very common culprit. How do engineers fight that?
• 17:23
  Fowling requires careful pretreatment of the feed water
• 17:26
  and periodic cleaning of the membranes for scaling,
• 17:30
  especially with calcium carbonate, engineers often add a
• 17:33
  small amount of acid to the feed water to lower the pH.
• 17:36
  This helps keep the carbonate dissolved instead of
• 17:39
  precipitating. But yeah, RO systems are high
• 17:41
  performance but demand careful operation and maintenance.
• 17:45
  OK, so pulling this all together, looking across both
• 17:47
  air and water, it seems like we keep seeing 2 main engineering
• 17:53
  philosophies at play. I think that's fair.
• 17:55
  On one hand you have prevention changing the process itself to
• 17:58
  avoid making the pollutant in the 1st place, like the staged
• 18:01
  mixing in Lonox burners. Right, process modification.
• 18:04
  And on the other hand you have a high efficiency separation or
• 18:06
  destruction after the pollutant is already formed like the SO2
• 18:10
  scrubbers or the DAC and tomorrow we just discussed for
• 18:14
  water. Yeah, the end of pipe treatment
• 18:16
  and increasingly the overarching goal, the philosophy that
• 18:19
  regulators and companies are pushing for is pollution
• 18:22
  prevention. P2.
• 18:23
  Why the big push for P2? Because at the end of the day,
• 18:27
  it's almost always more sustainable and often more cost
• 18:31
  effective in the long run to not generate the waste in the first
• 18:33
  place. If you can modify your process
• 18:36
  to avoid making SO2 or hazardous sludge, you save all the capital
• 18:40
  cost of building the scrubber or the sludge dryer, and all the
• 18:43
  operating costs for reagents and energy and disposal.
• 18:46
  Prevention beats cure. That makes a lot of sense, which
• 18:49
  leads us perfectly into our final thought.
• 18:51
  Kind of looking ahead, how do we optimize these incredibly
• 18:55
  complex control systems, running a power plant or a chemical
• 18:59
  facility to meet tight emission limits while still being
• 19:02
  efficient? It's tricky.
• 19:05
  It's incredibly tricky. These processes are dynamic
• 19:07
  inputs like fuel quality can change and the relationships
• 19:10
  like no X formation versus temperature are highly non
• 19:13
  linear. Simple controls aren't always
• 19:15
  enough. And our sources suggest this is
• 19:17
  where computer control, even artificial intelligence, is
• 19:20
  starting to play a bigger role. Definitely.
• 19:22
  We're moving beyond basic feedback loops.
• 19:25
  Think about optimizing combustion in a large boiler.
• 19:28
  You're juggling fuel flow, multiple air injection points,
• 19:31
  temperature sensors, Knox analyzers, CEO monitors.
• 19:34
  It's a complex dance. So how could?
• 19:36
  AI help an engineer manage that dance.
• 19:38
  What about methods like fuzzy logic or artificial neural
• 19:41
  networks? Okay fuzzy logic is interesting.
• 19:44
  It allows you to program control systems using more human like
• 19:47
  fuzzy rules based on expert knowledge.
• 19:49
  Instead of saying if temperature is exactly 900 C do X.
• 19:53
  You can say things like if Knox is high, any temperature is
• 19:56
  climbing then increase the over fire air slightly.
• 19:58
  So it handles ambiguity better, like how a human operator might.
• 20:01
  Think exactly It lets the automated system make nuanced
• 20:04
  adjustments based on qualitative states, not just precise
• 20:07
  numbers, which is often how real world systems behave.
• 20:10
  And artificial neural networks. ANSA.
• 20:13
  NNS are more about pattern recognition and learning from
• 20:15
  data. You could feed an A&N vast
• 20:18
  amounts of historical operating data, all the sensor readings
• 20:21
  that control settings, the resulting emissions.
• 20:24
  The network learns the complex, often non intuitive
• 20:26
  relationships between all those variables.
• 20:29
  Then, in real time, it can analyze the current state of the
• 20:32
  plant and predict which combination of control settings
• 20:35
  will keep production efficient while simultaneously minimizing
• 20:38
  pollutant emissions, keeping you within your permit limits.
• 20:41
  So it's like finding the operational sweet spot
• 20:43
  automatically, constantly adjusting based on learned
• 20:46
  patterns. That's the goal, An automated
• 20:48
  system that continuously optimizes that complex trade off
• 20:52
  between production efficiency, cost and environmental
• 20:55
  compliance. It's about turning that
• 20:57
  engineering challenge into a real time automated solution.
• 21:00
  A fascinating glimse into the future of process control and
• 21:03
  environmental engineering. Until the next equation design
• 21:06
  junkies, we are mechanical engineering made Simle.