How gas laws help understand DGA and predict problems before smoke appears (literally).

Dive into a world where gas tells the truth about the condition of multimillion-dollar investments. Discover the laws that are neither magic nor art—but pure physics.

If you work with transformer diagnostics, design substations, or manage energy infrastructure, understanding the basic gas laws can transform your approach to DGA—from intuitive to scientifically precise.

And that difference can save millions—not through "cost cutting" but through more accurate technical decisions.

Why are we talking about gas laws?

DGA (Dissolved Gas Analysis) is more than just “gut feeling and belief.” It’s the analysis of gases dissolved in transformer oil that can detect microscopic changes before a failure occurs.

But to truly understand what these gases are telling us, it’s worth starting with the physical laws that govern their behavior.

The ideal gas is not a myth. Even though reality is more complex, the ideal gas law equations provide a starting point for understanding diffusion, partial pressure, and equilibrium in the oil–gas system.

What exactly is Dissolved Gas Analysis (DGA)?

Dissolved Gas Analysis, or DGA, is a diagnostic method used in oil-immersed transformers. Its goal is to detect trace amounts of gases produced by thermal or electrical faults.

These gases dissolve in the insulating oil and serve as “fingerprints” of different types of degradation—before anything becomes visible to the naked eye.

Which gases are analyzed in DGA?

The most commonly monitored are seven key gases:

Hydrogen (H₂) – indicates early partial discharges and corona,

Carbon monoxide (CO;)
and carbon dioxide (CO₂) – linked to the degradation of insulating paper,

Methane (CH₄);
and ethane (C₂H₆) – signs of oil overheating,

Ethylene (C₂H₄) – higher temperatures, often associated with hot spots,

Acetylene (C₂H₂) – a marker of electrical arcing (the most dangerous type of fault).

What are the standards and gas tests?

ASTM D3612 is an international standard defining methods for extracting and measuring gases from transformer oil. It is complemented by standards like IEC 60567 and IEC 60599, which classify fault types based on gas ratios.

There is also frequent mention of the “three gas tests” in DGA:

• Gas ratio test (Rogers Ratio or Dornenburg) – comparing ratios of selected gases,

• Duval Triangle – a visual method for classifying faults based on three dominant gases,

• Threshold test – assessing whether the concentration of a specific gas exceeds defined alarm limits.

1. The ideal gas law – the foundation of it all

In the world of transformers, where precision can mean millions, the ideal gas law is not just a school formula—it is the foundation upon which the entire logic of Dissolved Gas Analysis (DGA) is built.

The state equation:

PV = nRT

can be treated as the mathematical DNA of gas behavior inside a transformer. And although a transformer is not a vacuum flask in a lab, its interior—especially the oil–gas system—operates according to the same physical principles.

What do the symbols mean?

P – gas pressure: how strongly the gas "pushes" against its surroundings.

In a transformer, this refers to the partial pressure of individual gases, either dissolved or above the oil surface.

V – the volume the gas occupies. Even when gases are dissolved in oil.

Their molar volume plays a role when estimating the amount of gas produced.

n – number of moles of gas.

This is key to understanding how much hydrogen, methane, acetylene, or carbon oxides were generated in a reaction.

R – the gas constant. Constant, yet not to be ignored.

A universal value that connects all variables into one logical framework.

T – temperature. Often non-uniform in transformers.

"Hot spots" can locally reach up to 200°C.

How does it work in practice?

Let’s assume a microscopic amount of acetylene forms due to a short circuit. Measuring its concentration in the oil is one thing. But only by knowing the temperature in the affected area and the pressure conditions can we calculate how much gas actually formed.

More importantly—does the amount indicate temporary overheating, or long-term degradation of cellulose?

The ideal gas equation lets us "go back in time"—drawing conclusions about causes based on the effects, i.e., the detected gases.

The transformer as a chemical reactor

Think of a transformer as a closed system, where every change in temperature or volume affects the state of gases.

Overheating increases T, which—if the volume is constant—increases P.

That’s why gas measurements must be correlated with temperature data. Without that, interpreting DGA would be like forecasting the weather by looking at clouds—too many unknowns.

2. Henry: how much does a gas “like” to dissolve?

Imagine a cold Coca-Cola straight from the fridge.

You open it and hear a hiss—that’s carbon dioxide escaping from the liquid. Now leave that same bottle in the sun. The result? The gas escapes faster, and the drink goes flat.

Exactly the same mechanism works in transformers. It’s governed by Henry’s law, one of the most underestimated yet essential phenomena in DGA interpretation.

What does Henry’s law say?

In its simplest form:

C = kH ⋅ P

C – concentration of gas dissolved in the liquid (mol/m³)
kH – Henry’s constant, depending on gas type and temperature
P – partial pressure of the gas above the liquid

In practice, this means that the higher the gas pressure, the more will dissolve in the oil. But! That’s only half the story—because Henry’s constant decreases with temperature, meaning the warmer it gets, the less gas can remain in the liquid.

How does this work in a transformer?

Imagine local overheating of cellulose insulation—CO and CO₂ are generated. These gases partly dissolve in oil and partly rise into the headspace. If the transformer’s temperature increases even slightly, the oil’s capacity to retain gas drops. As a result, more CO escapes into the “head,” and its concentration in the oil seemingly decreases—even though the degradation process may be intensifying.

Caution! This is a trap in interpretation. A lack of gas doesn’t always mean no damage—it might simply mean the gas has already escaped.

Every gas “prefers” something different

Different gases have different kH values:

• Hydrogen (H₂) – very poorly soluble, quickly escapes from oil

• Carbon dioxide (CO₂) – relatively soluble, “sticks around” longer

• Acetylene (C₂H₂) – short-lived, but detectable in arc faults

  Knowing these properties allows engineers to better assess whether a gas has just formed or if the sampling system recorded it with a delay.

Interpretation with physics in the background

In day-to-day DGA practice, it’s not only about knowing threshold values, but also understanding the physical context:

• Oil temperature – was it stable in recent days?

• Time since the event – did the gas have time to dissolve or separate?

• Do online readings differ from lab samples?

Henry’s law doesn’t give a ready-made answer, but it shows that gas isn’t just a number—it’s a physical phenomenon reacting to a dynamic environment. And that understanding builds an edge in transformer condition analysis.

3. What happens when temperature rises?

Temperature is not just the background to processes inside a transformer—it’s their primary catalyst. It determines whether chemical reactions ignite like a spark or remain dormant. For DGA interpretation, understanding the role of temperature is fundamental. It directly affects how many gases are formed, how quickly they move, and how long they remain dissolved in oil.

Heat as the trigger for gas formation

Inside the transformer, temperature conditions vary. Of critical importance are so-called hot spots—local points of elevated temperature, sometimes exceeding 200°C. This is where:

• Pyrolysis of cellulose insulation occurs (producing CO, CO₂)

• Thermal breakdown of oil takes place (producing CH₄, C₂H₆)

• Ethylene and acetylene form at extreme temperatures (above 500°C in arcing faults)

Rising temperature not only initiates gas-forming processes but also amplifies their intensity.

According to the Arrhenius equation:

k = A ⋅ e − Ea/RT

where:
k – reaction rate
A – frequency factor
Ea – activation energy
R – gas constant
T – temperature in Kelvin

The higher the temperature, the smaller the value of the exponential denominator, hence the faster the reaction. This means that even a slight increase in temperature (e.g., from 120 to 150°C) can exponentially accelerate gas production.

Temperature vs. gas solubility

High temperature not only creates gas—it also affects its behavior in oil. Back to Henry’s law: higher temperature means lower gas solubility in liquids. In practice, when the system heats up:

• More gas escapes from the oil to the headspace

• The dissolved gas concentration decreases—which may falsely suggest the “situation is improving”

• Partial pressure above the liquid increases—affecting further secondary reactions

Interpretation pitfalls

DGA performed while the transformer is operating (e.g., on a hot day) can yield different results than the same analysis done after cooling. That’s why each reading should be compared with temperature data: from online sensors, thermal history, or ideally—from hot spot temperature estimates (HST).

Without this, we risk a misinterpretation:

• Low gas concentration at high temperature does not necessarily mean no risk

• Sudden gas increase after cooling may reveal previously hidden processes

  ---

Relationships worth knowing

Effective DGA diagnostics requires knowing not only standards, but also physical interdependencies:

• Gas generation rate – increases exponentially with temperature

• Solubility – decreases with temperature

• Partial pressure – rises with temperature at constant volume

These three phenomena together create a dynamic system that cannot be understood solely through an alarm threshold table.

Only by accounting for the role of temperature can we see the full picture and anticipate possible fault development scenarios.

4. Dalton and the gas mixture

Unlike in a laboratory, inside a transformer we never deal with just one gas. Degradation processes produce a whole spectrum of compounds—from light hydrogen to complex hydrocarbons.

That’s why, instead of analyzing each gas in isolation, it’s important to understand how they behave collectively. Here, Dalton’s law becomes one of the key gas laws in the context of DGA.

What does Dalton’s law say?

Ptotal = P1 + P2+ ⋯ + Pn

This means that the total pressure of the gas above a liquid (such as in the transformer headspace) is the sum of the partial pressures of all its components.

Each gas contributes its “share” to the total pressure—proportional to the number of moles present in the mixture.

Why is this important? Because in a transformer, it’s this very gas mixture—and its changing proportions—that reveals the type and intensity of the fault.

The mixture as a fault fingerprint

By analyzing the gas mixture composition, we can identify dominant degradation mechanisms:

• A predominance of hydrogen (H₂) and methane (CH₄) suggests partial discharges,

• The presence of acetylene (C₂H₂) is a clear sign of arcing,

• High levels of CO and CO₂ indicate cellulose paper insulation degradation,

• Increased ethylene (C₂H₄) is typical for overheating.

Dalton’s law allows us to model how partial pressures vary over time.

This in turn helps detect whether any particular gas is increasing rapidly—potentially indicating an escalation of the fault before it becomes apparent in summary charts..

Gas escape dynamics

Each gas in the mixture has a different solubility coefficient (see Henry’s law), but Dalton’s law determines which gas escapes the liquid first.

Those with higher partial pressures (e.g., hydrogen) will reach equilibrium between the oil and gas phases faster—and disappear from the system more quickly.

This explains why laboratory samples don’t always reflect the full spectrum of gases that were present moments earlier.

The absence of a gas in the sample doesn’t necessarily mean it’s no longer present in the transformer—it may simply have diffused or been vented out earlier.

IInterpreting gas ratio changes

In practice, diagnostics often rely on gas ratio tests, such as the Dornenburg or Rogers methods. It is thanks to Dalton’s law that these methods make sense: they allow us to evaluate not only how much gas formed, but how the various components relate to one another.

A noticeable shift in the ratio of, say, C₂H₂ to CH₄ may indicate a change in the fault type—e.g., from overheating to arcing.

If, on the other hand, gas ratios remain stable while concentrations increase evenly—this suggests the same fault is simply progressing.

Practical conclusions

• Don’t analyze gases in isolation—the context of the mixture matters,

• Watch for ratio changes—they're more revealing than absolute values,

• If a gas disappears from the sample—check the pressure, temperature, and sampling history. It may have simply left the system.

Dalton’s law offers a holistic view of the gas system—not just as individual indicators, but as a dynamic system where every change has causes and consequences.

5. Diffusion – gas never sleeps

Gases in a transformer are not passive indicators of faults. They are active, mobile particles that—even after the fault processes stop—continue to “live their own lives”—slowly spreading through the system, reaching equilibrium, vanishing from samples or appearing where they weren’t before. This is governed by diffusion, precisely described by Fick’s first law.

What does Fick’s law say?

J = −D ⋅ dc/dx

Where:
J – diffusive flux (amount of moles moving through a surface per time unit),
D – diffusion coefficient (specific for each gas and medium),
dc/dx – concentration gradient (difference in gas concentration across space)

In short: gas moves from where there's more of it to where there's less—and the greater the difference, the faster the movement.

What does this mean in practice?

There is no such thing as a “constant gas composition” in a transformer—especially in systems with a large oil volume. Even if the fault occurs in a single spot (e.g., a local short), the generated gases will slowly spread throughout the entire system.

If a sample is taken from a different location than the fault origin—the results may be underestimated.

If analysis is delayed—the gas may have already escaped or diffused, blurring the alarm signal.

The importance of time – DGA isn’t always real-time

What we measure in a sample is a snapshot of the system at that moment. But diffusion means the system is constantly changing—even after the gas-forming reactions have ceased. In practice, this leads to several key recommendations:

• A measurement taken immediately after the fault gives a different profile than one taken a week later,

• The smaller the transformer, the faster diffusion equalizes concentrations,

• Online systems allow for dynamic tracking—classic lab analysis shows only the “averaged effect.”

Why does diffusion matter for interpretation?

Imagine a transformer where ethylene (C₂H₄) was generated due to overheating. As soon as the temperature drops, the gas-forming process stops—but the ethylene continues to move through the oil. If sampling is delayed, the gas will already be partially dispersed or even vented into the headspace.

The result? The measurement shows a lower concentration than what actually existed at the moment of the fault.

The same goes for hydrogen—very light, poorly soluble, and prone to rapid diffusion. If the measurement is not taken in time, hydrogen may be incorrectly interpreted as absent—even though it was one of the first fault indicators.

Practical conclusions

• Interpret DGA considering the time and location of the sample,

• Use online systems wherever possible—they give a more complete picture of the dynamics,

• Understand that the absence of gas doesn’t always mean no issue—it might be the result of diffusion or escape.

Fick’s law helps us better understand how the system “cleans itself” of gases—and how quickly fault information can fade.

It’s physics at work—continuously—even when everything seems to have returned to normal.

Let’s interpret the data that matters

In a world where the speed of decision-making matters more than the number of decisions made, access to reliable data becomes one of the most important advantages. But data alone is not enough.

Only proper interpretation—based on physics, process understanding, and real-world experience—creates value that allows us to protect, optimize, and plan the future of power infrastructure.

That’s why today, instead of asking whether DGA “shows something,” we ask: what exactly does it show, and how can we act smarter because of it?

At Energeks, we believe that every network device—from transformers to energy storage—deserves the same level of precision as the most advanced IT systems. Diagnostics doesn't have to be a guessing game—it can be science-based, predictable, and transparent. And that’s precisely what understanding gas laws enables.

As one of Europe’s leading suppliers of medium-voltage transformers and transformer stations, we support our clients daily in making decisions with long-term technical, financial, and environmental consequences.

That’s why our portfolio continues to grow:
➤ Modern transformers and complete transformer substations
➤ Energy storage systems, inverters, and EV charging infrastructure
➤ Technologies for photovoltaic farms and the renewable energy sector—efficient, safe, and future-ready

We proudly support investors, designers, municipalities, and technology integrators in creating solutions that work not only today—but also tomorrow.

Technology is the tool. People and values are the direction.

Get in touch with us if you’d like to discuss your challenge—we’re here to share our experience and find the best solutions together.

And if you’d like to become part of our knowledge and inspiration network—join us:
➤ Connect with the Energeks community on LinkedIn

Thank you for being with us—together we are building an infrastructure that not only works, but… learns, adapts, and grows alongside you.

Source:
Transformers Magazine vol.12