Electronics Guide

Types of Corrosive Settings

Industrial corrosion arises in several characteristic environments. Marine and coastal settings combine chloride-rich salt spray with high humidity and dissolved oxygen, an aggressive mixture that promotes rapid rusting of steel and pitting of many alloys. Chemical process environments expose equipment to acids, caustic alkalis, oxidizers, and organic solvents, sometimes at elevated temperature and pressure that accelerate attack. Oil and gas production introduces hydrogen sulfide and carbon dioxide, which dissolve in water to form acidic, embrittling conditions known as sour service, governed for metallic materials by ANSI/NACE MR0175 / ISO 15156. Wastewater and other biologically active settings generate hydrogen sulfide and its oxidation product sulfuric acid, and they support microbially influenced corrosion. Each of these environments demands its own material choices, since an alloy or coating that resists one chemistry may fail quickly in another.

The physical form of the exposure matters as much as the chemistry. Continuous immersion, alternating wet and dry cycles, condensing vapors, splash zones, and aerosol mists each create different corrosion patterns. The splash zone of a marine structure, repeatedly wetted and richly aerated, is typically the most corrosive region of all: published field and laboratory studies place uncoated carbon steel there in the range of roughly 0.3 to 0.6 millimeter of thickness loss per year, with reported peaks approaching 1 millimeter per year, well above the rate of the same steel fully submerged. Temperature swings drive condensation that concentrates corrosive species on cool surfaces. A harvester placed in such a setting experiences not a single steady condition but a fluctuating combination of chemical, thermal, and humidity stresses that together determine how long it survives.

Classifying Environmental Severity

Because "corrosive" spans an enormous range of severity, standardized classifications let designers state a requirement precisely instead of arguing about adjectives. For outdoor atmospheres, ISO 9223 defines corrosivity categories C1 (very low) through C5 (very high) plus CX (extreme), each anchored to the measured first-year corrosion rate of standard metal coupons. Category CX, which covers severe marine and marine-industrial atmospheres such as offshore platforms and coastal chemical plants, corresponds to a first-year loss of roughly 200 to 700 micrometers on carbon steel. Specifying a target category converts a vague environmental description into a testable design input and points directly at the coating systems, defined in the companion standard ISO 12944 for protective paint systems on steel structures, that are qualified for it.

For the electronics inside the harvester, the relevant classification concerns airborne contaminants rather than bulk metal loss. ANSI/ISA-71.04 grades gaseous severity by exposing copper and silver coupons for thirty days and measuring the thickness of corrosion product formed: G1 (mild) below about 300 angstroms of copper reaction, G2 (moderate) below about 1,000, G3 (harsh) below about 2,000, and GX (severe) above that. Refineries, pulp mills, wastewater plants, and fertilizer facilities routinely reach G3 or GX, levels at which unprotected circuit assemblies fail in months. Stating both an ISO 9223 category for the housing and an ISA-71.04 severity level for the interior gives a corrosion-resistant harvester two clear, independently verifiable targets.

Corrosion Mechanisms Affecting Devices

Several distinct mechanisms threaten a harvester and its electronics. Uniform corrosion thins exposed metal at a roughly predictable rate. Localized attack, including pitting and crevice corrosion, concentrates damage at small sites and can perforate a wall while most of the surface remains intact, making it especially dangerous because it is hard to detect and progresses quickly. Galvanic corrosion accelerates attack where dissimilar metals contact in the presence of an electrolyte, a frequent hazard in devices that combine several materials. Stress corrosion cracking and hydrogen embrittlement can cause sudden brittle failure of loaded components in specific chemical environments, often without prior warning.

For electronics, the most insidious threat is the ingress of moisture and ions. Even trace humidity penetrating a package can corrode metallizations, bridge conductors with conductive films, and cause leakage currents that disrupt sensitive low-power circuits. Where a voltage bias, an ionic residue, and adsorbed moisture coexist, electrochemical migration grows conductive dendrites between adjacent conductors and shorts them. Corrosive gases such as hydrogen sulfide attack copper and silver, tarnishing contacts and degrading connections; in sulfur-rich atmospheres, creep corrosion spreads copper or silver sulfide across a board surface and bridges fine-pitch features that the gas never directly attacked. Because harvesting electronics often operate at very low power and depend on tiny currents and high-impedance nodes, they are unusually sensitive to the leakage and contamination that corrosion produces, which makes the integrity of their enclosure paramount.

Thermoelectric Harvesting from Process Heat

Chemical and petrochemical plants abound in temperature differences. Reactors, distillation columns, heat exchangers, steam lines, and hot product streams run well above ambient, and the gradient between a hot pipe surface and the surrounding air is a dependable, continuous energy source. Thermoelectric generators clamped to such surfaces convert this gradient directly into electricity with no moving parts, an advantage in settings where mechanical complexity and maintenance are undesirable. Because the heat is a byproduct of the process and flows continuously, thermoelectric harvesting offers a steady baseline supply well matched to the needs of always-on monitoring sensors.

In a corrosive setting the thermoelectric module itself must be shielded from the atmosphere, since its bismuth telluride legs, solder joints, and electrical interconnections corrode readily if exposed. The usual approach mounts the module within a sealed, corrosion-resistant housing that conducts heat from the pipe to the module hot side through a protected thermal path while rejecting heat to the air through a resistant heat sink. The hot-side interface to the process surface, the cold-side heat exchanger, and the enclosure must all withstand the local chemistry.

The design is dominated by thermal resistance rather than by the module itself. Only a fraction of the pipe-to-air temperature difference appears across the thermoelectric legs; the rest is lost in the mounting interface, the sealed thermal path through the housing wall, and above all the convective resistance of the heat sink, which corrosion-resistant coatings and fouling only worsen. A commercial pipe-clamp harvester therefore delivers from a few milliwatts to a few hundred milliwatts, scaling with module area and with the temperature difference actually achieved. That range is generous next to the demand of a wireless process transmitter, which averages a fraction of a milliwatt when it reports at intervals of seconds to minutes, so the practical design target is a sufficient cold-side heat sink rather than a more efficient module. With these protections, thermoelectric harvesting is one of the most practical and widely applicable methods for powering sensors on hot industrial equipment.

Vibration and Flow Harvesting

Rotating machinery is ubiquitous in process plants and on marine structures, and pumps, compressors, fans, and motors transmit continuous vibration to the structures that support them. Piezoelectric and electromagnetic vibration harvesters convert this oscillation into electricity, conveniently co-locating power generation with the very machinery whose health is being monitored. Useful excitation clusters at the running speed and its harmonics and at the line frequency of motors, so a resonant harvester must be tuned to a frequency the machine actually produces; a device tuned even a few percent away from the excitation loses most of its output. Flow within pipes offers another source: turbulent flow, vortex shedding past an obstruction, and the bulk movement of liquids and gases can drive small turbines or excite vibrating harvesters, allowing energy to be drawn from the process stream itself, at the cost of a small permanent pressure drop that the process must tolerate.

Both vibration and flow harvesting demand careful sealing in corrosive service because moving parts and exposed transducers are vulnerable. Flow harvesters placed in the process fluid contact the most aggressive chemistry directly and must be built from resistant materials or protected by coatings; they also become a fouling and blockage risk in dirty or scaling streams. Vibration harvesters can often be mounted externally in a sealed housing, contacting only the structure rather than the process fluid, which simplifies their protection considerably and is the preferred arrangement wherever the vibration is strong enough. Where moving components such as turbine rotors, bearings, or suspended magnetic masses are unavoidable, they are made from corrosion-resistant alloys, run on magnetic or fluid-film supports, or are separated from the fluid by a non-magnetic containment can that lets torque cross the barrier without a rotating seal.

Radio-Frequency and Solar Harvesting

Industrial sites carry radio-frequency energy from wireless communications, control systems, and broadcast sources, and dedicated transmitters can also be deployed to power nearby sensors deliberately. Rectifying antennas capture this ambient or intentional radio energy and convert it to direct current. Because the antenna and rectifier can be fully encapsulated behind a chemically resistant but radio-transparent window of fluoropolymer, engineering thermoplastic, or ceramic, radio-frequency harvesting is well suited to corrosive environments where any opening or moving part would be a liability. Ambient field strengths at a typical plant yield only microwatts over practical antenna areas, however, so ambient harvesting suits sensors that wake for a few milliseconds every several minutes, and a deliberately placed transmitter is usually needed to power anything more demanding.

Where corrosive equipment stands outdoors or under lighting, photovoltaic harvesting supplements other sources. Solar cells protected behind a corrosion-resistant transparent cover can power outdoor sensors on tank farms, pipelines, and offshore structures. The protective glazing must resist both the chemical atmosphere and ultraviolet degradation while remaining transparent, which favors low-iron tempered glass or fluoropolymer front sheets over ordinary acrylic. Fouling is the practical limit rather than corrosion: salt crust, dust, and process deposits on the window reduce output steadily, so panels are oversized against soiling and mounted at an angle that sheds contamination. As with radio-frequency harvesting, the appeal of solar in these settings is that the energy-capturing surface can be fully sealed behind a passive window, leaving no exposed conductor or moving part for the chemistry to attack.

Electrochemical and Corrosion-Cell Harvesting

A distinctive option in corrosive service is to draw power from the electrochemistry itself. Two dissimilar electrodes immersed in a conductive electrolyte, whether seawater, produced water, or wastewater, form a galvanic cell that delivers a small continuous current as the less noble electrode consumes itself. Seawater-activated cells and sacrificial-anode couples exploit this directly, and a sensor may also tap the existing cathodic-protection circuit of a pipeline or platform, which already circulates current through the structure and its surroundings. The energy is real but bounded by the anode mass, so such a harvester is better understood as a very long-lived primary battery whose fuel is the corrosion the structure was going to suffer anyway.

Biologically active environments offer a related mechanism. Microbial fuel cells oxidize organic matter in sediment or wastewater at an anode colonized by electroactive bacteria, delivering power densities on the order of tens of milliwatts per square meter of electrode. That is far too little for continuous transmission, but with a supercapacitor buffer and a duty cycle measured in minutes it suffices for water-quality and corrosion sensors in the very tanks and channels whose chemistry makes conventional powering difficult. Both approaches share an important caveat: the harvester is now electrically coupled to the structure, and the designer must confirm that the current it draws does not disturb the cathodic protection it borrows from or create a new galvanic cell of its own.

Corrosion-Resistant Alloys and Polymers

Structural and housing components of corrosion-resistant harvesters are commonly built from alloys selected for the particular environment. Austenitic stainless steels such as type 316L resist many atmospheres but suffer pitting and crevice corrosion in chloride-rich marine and process conditions, where duplex grades, six-percent-molybdenum superaustenitic steels, or nickel-based alloys offer greater resistance. Engineers compare candidates using the pitting resistance equivalent number, conventionally chromium content plus 3.3 times molybdenum plus 16 times nitrogen, all in weight percent: type 316L sits near 25, duplex 2205 near 35, and superaustenitic and nickel-chromium-molybdenum alloys above 40. Titanium provides outstanding resistance to seawater and to oxidizing chemicals and is favored for demanding marine service, although commercially pure grades can still suffer crevice corrosion in hot, concentrated chloride brines, where palladium-bearing grades are specified instead. The choice always depends on the specific chemistry: an alloy excellent in one acid may corrode rapidly in another, and molybdenum-bearing alloys that excel against chlorides can be poor choices in strongly oxidizing acids.

Where metals are unsuitable, engineering polymers and composites provide an alternative. Fluoropolymers such as polytetrafluoroethylene resist nearly all chemicals and make excellent barriers and seals. Other resistant plastics, including polyether ether ketone and polyphenylene sulfide, along with fiber-reinforced composites and technical ceramics, serve as housings, windows, and structural elements that no electrolyte can corrode in the way it attacks metal. Non-metallic construction also sidesteps galvanic corrosion entirely, which is a significant advantage in devices that would otherwise combine several dissimilar metals.

Polymers do not fail by corrosion, but they fail nonetheless, and the mechanisms deserve explicit attention. Every polymer is permeable to water vapor and to gases, so a plastic wall slows ingress rather than stopping it; polytetrafluoroethylene in particular combines superb chemical resistance with relatively high gas permeability and a marked tendency to cold flow under sustained gasket load. Solvents swell and soften many thermoplastics, and combined tensile stress and a specific chemical agent can produce environmental stress cracking in components that would tolerate either alone. Elastomeric seals in high-pressure gas service can be destroyed by rapid gas decompression, when absorbed gas expands within the seal as pressure falls and tears it internally. Polymer construction also brings lower strength, greater thermal expansion, and far lower thermal conductivity, the last of which directly penalizes any thermoelectric design that must pass heat through the wall.

Protective Coatings and Surface Treatments

Coatings extend the life of components that cannot be made entirely of resistant material. Organic coatings such as epoxies, polyurethanes, and fluoropolymer films form a barrier between metal and environment and are widely used on housings and external surfaces; ISO 12944 codifies the multi-coat systems, dry-film thicknesses, and surface preparation grades appropriate to each corrosivity category and to a required durability range. Inorganic and metallic coatings, including electroplated, hot-dip, and thermally sprayed layers, protect by forming a resistant surface or by sacrificing themselves galvanically to spare the underlying metal; thermally sprayed aluminum is a common choice on offshore steel for exactly this reason. Conformal coatings of parylene, acrylic, silicone, or urethane are applied directly to circuit boards to shield the electronics from moisture and contamination, an important second line of defense behind the enclosure.

Conformal coating deserves a realistic assessment, because it is often asked to do more than it can. Vapor-deposited parylene C, applied a few micrometers to a few tens of micrometers thick, conforms to sharp edges and covers surfaces that a sprayed coating cannot reach, and it is the usual choice for the harshest gaseous environments. No conformal coating is hermetic, however: all of them pass moisture vapor slowly, and any coating applied over an unclean board seals ionic residue in against the metal rather than out. Effective practice therefore treats coating as one layer of a system that begins with a clean, low-residue assembly process, and it accepts that connectors, test points, and rework sites are the places where coverage fails.

Surface treatments alter the metal itself to improve resistance. Passivation of stainless steel dissolves free iron from the machined surface and restores its protective chromium oxide film. Anodizing thickens the natural oxide on aluminum and titanium, and a sealed hard anodic layer substantially improves the corrosion resistance of an aluminum housing. The effectiveness of any coating depends critically on its integrity, because a single defect, scratch, or pinhole concentrates attack over a small anodic area and can drive rapid undercutting that lifts the coating from beneath. Coatings are therefore chosen and applied with attention to surface preparation, adhesion, thickness, and freedom from defects, and they are often combined in layers so that the failure of one does not immediately expose the substrate.

Enclosure Sealing Strategies

Protection of harvesting electronics ranges from environmental sealing to true hermeticity. Sealed enclosures using elastomeric gaskets and rated ingress protection keep out liquid water and dust and suffice for many moderately corrosive settings; IEC 60529 codes such as IP66 for powerful water jets and IP68 for continuous immersion describe what these enclosures achieve. The rating is a liquid-water test, however, and says nothing about vapor, so polymer seals rated IP68 still admit moisture vapor steadily over years. Truly hermetic packages, closed by welding, brazing, or glass-to-metal seals, exclude even water vapor; their integrity is verified by leak testing to procedures such as MIL-STD-883 Method 1014, which pairs a helium fine-leak measurement with typical acceptance limits in the range of 10⁻⁸ to 10⁻⁷ standard cubic centimeters per second against a gross-leak test that catches large defects. The level of sealing is matched to the severity of the chemistry and the required lifetime, since hermetic construction is markedly more costly but offers the only reliable protection against the slow permeation that eventually defeats polymer seals.

A sealed enclosure that is not hermetic must also cope with pressure differences. Thermal cycling and weather make a closed cavity breathe through its weakest point, drawing corrosive liquid past a gasket in a mechanism sometimes called seal pumping. Enclosures in such service are fitted with a vent membrane of expanded polytetrafluoroethylene that equalizes pressure while blocking liquid water, or they are made genuinely hermetic so that no pressure-driven exchange can occur. Gasket materials must themselves suit the chemistry: fluoroelastomers cover many acids and hydrocarbons, perfluoroelastomers extend that range at high cost, and ethylene propylene diene monomer handles steam and many alkalis but degrades in hydrocarbons.

Feedthroughs are the critical weak point of any sealed harvester, because energy and signals must cross the barrier while the barrier remains intact. Thermoelectric harvesters must pass heat through the wall, vibration harvesters must couple mechanical motion, and every device must bring electrical connections out to sensors. Glass-to-metal and ceramic-to-metal feedthroughs provide hermetic electrical penetrations, and their sealing metals must be chosen for thermal expansion compatibility as well as for corrosion resistance, since a cracked glass seal fails completely rather than gradually. Thermal coupling is achieved through sealed conductive paths that inevitably cost temperature difference, and mechanical coupling through sealed flexible mounts or non-contact magnetic links that transmit force with no penetration at all. Designing these penetrations to transmit the necessary energy without compromising the seal is among the central challenges of corrosion-resistant harvester engineering.

Potting, Encapsulation, and Desiccants

Filling the interior of a device with a protective compound provides robust protection where a sealed cavity is unnecessary or impractical. Potting and encapsulation embed the electronics in epoxy, silicone, or polyurethane, excluding air and moisture and supporting the components mechanically against vibration and shock. A solid encapsulant leaves no void for condensation to form and can resist pressure as well as chemistry, which suits subsea and high-pressure corrosive applications. The encapsulant must be chosen for compatibility with the components and for stability in the operating temperature range, since some compounds become brittle or shrink with age and thermal cycling.

Where a sealed air-filled cavity is used, residual and permeated moisture is managed with desiccants and by controlling the internal atmosphere. Sealing the enclosure with dry gas or under partial vacuum, and including a desiccant to absorb any moisture that enters, prevents internal condensation that would otherwise corrode the electronics from within. Getter materials can absorb specific corrosive gases. These measures recognize that perfect sealing is difficult to achieve and maintain over years, so a well-designed corrosion-resistant device combines barriers that exclude most contamination with internal measures that neutralize the small amount that inevitably penetrates.

Galvanic Compatibility and Integrated Protection

Because galvanic corrosion accelerates attack wherever dissimilar metals meet in an electrolyte, the selection and arrangement of materials within a device must be considered as a whole. Combining metals that are widely separated in the galvanic series invites rapid corrosion of the less noble metal at their junction. Designers minimize this risk by selecting compatible materials, by isolating dissimilar metals with insulating barriers, and by avoiding electrolyte pathways that would complete a galvanic circuit. In some installations a sacrificial anode or an impressed-current system deliberately protects critical components by corroding in their place, a technique borrowed from the broader practice of cathodic protection.

Protection is most effective when designed into the device rather than added afterward. A harvester intended for corrosive service is conceived from the start with sealed construction, resistant materials, compatible material pairings, and protected feedthroughs as integral features. Drainage paths that prevent the pooling of corrosive liquid, smooth geometries that shed condensation, and the elimination of crevices where electrolyte could stagnate all reduce corrosion before it begins. This integrated approach yields a device whose resistance does not depend on any single barrier but on the consistent application of resistant design throughout.

Energy Storage and Power Management in Harsh Service

Harvested energy must be conditioned and buffered, and the storage and power-management subsystem faces the same environmental stresses as the harvester. Supercapacitors and rechargeable cells smooth the intermittency of harvesting and supply the bursts of power that wireless transmissions require, since a radio drawing tens of milliamperes for a few milliseconds demands far more instantaneous current than a thermoelectric or radio-frequency source can provide. These storage elements and their associated electronics must be sealed and protected as carefully as the transducer. Temperature is the dominant life factor: the aging of both supercapacitors and lithium cells accelerates sharply with temperature, so a harvester clamped to a hot process line must either place its storage on the cool side of the assembly or adopt a chemistry rated for the service temperature, and lithium thionyl chloride or high-temperature supercapacitors are common answers.

Power-management electronics condition the harvested input into a regulated supply, track the maximum power point of the source, and govern the duty cycle of the sensor to match consumption to available energy. In corrosive service these circuits are especially sensitive to leakage caused by any moisture that penetrates the package: a surface leakage path of a few hundred megohms is irrelevant in a mains-powered instrument but ruinous at a cold-start threshold of a few hundred millivolts and a quiescent budget of a few hundred nanoamperes. Clean assembly, generous conductor spacing, guard rings around high-impedance nodes, and conformal coating are therefore functional requirements rather than cosmetic ones. Robust power management also allows the device to ride through periods when the harvester cannot meet demand, drawing on stored energy and shedding noncritical activity, which is valuable in a setting where a maintenance visit to replace a failed unit is difficult and costly.

Reliability, Testing, and Service Life

Demonstrating that a device will survive its intended service life in a corrosive environment requires deliberate testing, because field failures are expensive and sometimes hazardous. The most familiar procedure is the neutral salt spray test of ASTM B117 and ISO 9227, which holds specimens in a continuous fog of five percent sodium chloride at 35 degrees Celsius at a pH between 6.5 and 7.2. Its value and its limits are both well established: it ranks bare metals and platings usefully, but its continuous wetness suppresses the wet-and-dry cycling that drives real atmospheric corrosion, and its correlation with field life is poor for painted and coated parts. Cyclic corrosion tests such as ASTM G85 and ISO 11997-1, which alternate salt fog, dry-off, humidity, and in some variants acidified or sulfur-bearing atmospheres, reproduce field degradation far more faithfully and are the better basis for a life claim.

Other exposures target other mechanisms. Damp-heat and temperature-humidity cycling to IEC 60068-2-30 and IEC 60068-2-38 drive moisture through seals and provoke condensation inside enclosures. Flowing mixed gas testing to IEC 60068-2-60, using controlled concentrations of hydrogen sulfide, sulfur dioxide, nitrogen dioxide, and chlorine, reproduces the gaseous attack that ISA-71.04 classifies and is the appropriate test for a harvester destined for a refinery or wastewater plant. Immersion in the actual process chemistry, at the actual temperature, remains the final arbiter for any wetted part. Because acceleration factors between any of these tests and real service are uncertain, prudent programs pair laboratory qualification with instrumented field trials and with witness coupons installed alongside the first production units.

Designing for the full service life means anticipating the gradual nature of corrosion and the slow permeation of seals. Redundancy in protection, generous corrosion allowances on structural members, and the selection of materials with proven long-term performance all contribute to reliability. Because some degradation is inevitable, the most robust designs fail gracefully and, where possible, report their own deteriorating condition, through an internal humidity sensor or a monitored seal, so that they can be replaced on a planned basis rather than failing unexpectedly. The high cost and difficulty of access in corrosive installations make this emphasis on demonstrated, long-term reliability central to the value of self-powered monitoring.

Chemical Plant and Refinery Monitoring

Chemical processing plants and refineries deploy large numbers of sensors to track temperature, pressure, flow, and the integrity of vessels and piping. Self-powered sensors harvesting energy from the abundant process heat and machinery vibration eliminate the cost and ignition concerns of running power cables through hazardous areas and remove the need to service batteries on equipment that may be difficult or dangerous to reach. Corrosion-monitoring sensors are especially valuable, since the early detection of wall thinning, pitting, or coating failure allows maintenance to be scheduled before a leak or rupture occurs.

Because many areas of a chemical plant are also classified as explosive atmospheres, harvesting devices for these locations must combine corrosion resistance with intrinsically safe or explosion-proof construction under the IEC 60079 series, whether by limiting stored and released energy in an intrinsically safe design, by containing an internal explosion in a flameproof enclosure, or by encapsulating the circuitry in compound. The two requirements interact: an intrinsic safety assessment counts every capacitor and inductor in the harvester, so the storage capacitance that buffers a radio burst is precisely what constrains the certification, and the encapsulation that satisfies one protection concept also serves the corrosion barrier. A sensor that is both sealed against the corrosive atmosphere and certified for its hazardous-area zone can be placed where conventional powered instrumentation would be impractical, extending monitoring coverage into the most demanding regions of the plant and improving both safety and process efficiency.

Marine and Offshore Corrosion Monitoring

Marine structures, ships, offshore platforms, and subsea equipment suffer relentless attack from seawater, salt spray, and humid, oxygen-rich air, and corrosion is a primary determinant of their service life. Self-powered sensors monitoring corrosion rate, coating condition, cathodic protection status, and structural integrity allow operators to manage deterioration proactively across structures that are large, remote, and costly to inspect. Harvesting from wave motion, from the vibration of machinery and the structure itself, from thermal gradients, and from solar exposure provides power without batteries that would be impractical to replace on a remote platform or an underwater installation.

Pipelines, storage tanks, and water-treatment facilities present similar opportunities on land. Buried and submerged pipelines benefit from distributed corrosion sensors powered by harvesting from product temperature, flow, or the cathodic-protection currents already applied to the line. Wastewater treatment plants, with their corrosive gases and biologically active fluids, use sealed self-powered sensors to monitor both the process and the corrosion of their own infrastructure. In all of these settings, the combination of a chemically resistant enclosure with an energy source drawn from the environment allows continuous monitoring of assets whose failure would be expensive, polluting, or dangerous.