Capacitors Insulation Resistance Leakage Current and BDV

Capacitor leakage current is the small current that continues to flow under an applied DC voltage after charging and absorption transients have diminished. Insulation resistance expresses the same leakage behaviour as a resistance at specified test conditions. This article explains their relationship, measurement, operating dependencies and implications for capacitor selection by design, component and purchasing engineers.

Key Takeaways

Insulation resistance and DC leakage current are related by the voltage across the capacitor divided by the measured current; resistance and current are not direct mathematical reciprocals.
A meaningful leakage or insulation-resistance specification includes test voltage, temperature, measurement time and any required preconditioning.
Initial charging current and dielectric absorption current must be distinguished from the longer-term leakage current.
Ceramic and film capacitors commonly use insulation-resistance specifications, while electrolytic capacitors commonly specify leakage current. Compare individual datasheets rather than assuming a universal ranking.
Leakage affects charge retention, standby consumption and errors in high-impedance circuits. In power converters, it must be considered alongside capacitance, ESR, ripple current, temperature and lifetime.
Leakage alone does not establish reliability across different technologies. Out-of-limit leakage, instability or changes after stress require investigation under controlled test conditions.

Charge Retention and Voltage Ratings

Capacitance describes charge storage, but it does not completely describe a capacitor’s practical behaviour. Charge retention and permissible voltage are separate selection requirements.

Charge retention: internal conduction and surface leakage allow a charged capacitor to discharge even when no intentional external load is connected. Insulation resistance, abbreviated IR, or DC leakage current, abbreviated DCL, characterises this behaviour under defined conditions.
Voltage capability: rated voltage defines permissible operation under the manufacturer’s stated conditions. It is not the same as dielectric breakdown voltage, and it does not remove restrictions on temperature, polarity, ripple current or transients. A category voltage, where specified, applies under its stated temperature conditions.

Electrostatic capacitors, including ceramic, paper and organic film types, are commonly specified by IR. Aluminium and tantalum electrolytic capacitors are commonly specified by DCL. These conventions reflect their construction and customary specification practice, not different physical definitions of leakage. Rated, surge and forming voltages must also remain distinct; forming voltage is not an additional operating allowance. [1, 2]

Basic Charge Storage and Discharge

Charging and discharging rates depend on the complete circuit, not on capacitance alone. A simple resistor-capacitor circuit gives an exponential response when the source is a voltage step and resistance and capacitance can be treated as constant. Real sources, wiring, capacitor ESR and inductance limit the initial current and may alter the response. Connecting a capacitor directly to a low-impedance source, or short-circuiting a charged capacitor, can produce a damaging current pulse and is not a recommended test method.

Original equation image for capacitor charging through a resistance — Figure 1. Original capacitor charging equation. Source: Knowles Precision Devices.

Original equation image for capacitor discharge through a resistance — Figure 2. Original capacitor discharging equation. Source: Knowles Precision Devices.

Figure 3 compares the charging and discharging circuits and their voltage responses. Oscillators, waveform shapers and backup circuits exploit controlled changes in stored charge, although practical backup time also depends on load current and the minimum acceptable supply voltage. [3]

Resistor-capacitor charging and discharging circuits with their voltage curves — Figure 3. Capacitor charging circuit and voltage response, top, compared with the discharging circuit and voltage response, bottom. Source: Knowles Precision Devices. [3]

Insulation Resistance IR and Capacitor Leakage Current DCL

A real dielectric is not a perfect insulator. Its leakage can be represented by a parallel resistance for a specified operating point, but that resistance need not remain constant with voltage, temperature or time.

Capacitor equivalent circuit with insulation resistance connected in parallel with capacitance — Figure 4. Parallel-resistance representation of a capacitor’s insulation resistance.

The large electrode area and small dielectric thickness that provide useful capacitance also influence leakage. Figure 4 shows the parallel-resistance model. Terminal measurements can additionally include leakage over the package, between leads and through the test fixture or printed circuit board.

Measurement of IR and Leakage Current

To determine IR, apply the specified DC test voltage and measure the resulting current after the specified electrification time. The charging circuit contains source resistance, any intentional current-limiting resistance and the measurement instrument’s contribution. These affect charging time and can cause the voltage across the capacitor to differ from the source setting. [1, 4]

Capacitor charging circuit and exponential voltage rise through a series resistance — Figure 5. Charging circuit and voltage response with series resistance.

Figure 5 illustrates the charging interval. The current initially includes the current required to establish the capacitor voltage. Dielectric polarisation and other absorption processes can then produce a slower-decaying current. Consequently, the first current reading after voltage application is not a direct measurement of steady leakage. [1, 4]

Comparison of simplified capacitor charging current and a slower current decay including polarisation — Figure 6. Simplified charging-current response compared with a response including dielectric polarisation.

The simplified current curve in Figure 6 should not be interpreted as a perfectly ideal capacitor having finite leakage. A perfectly ideal capacitor has no steady DC leakage. A practical simplified model combines ideal capacitance with a finite parallel resistance; absorption requires additional time-dependent elements. For the capacitor itself, the measured current can be separated conceptually into charging current I_charge, absorption current I_abs and leakage current I_leak. Equation (1) expresses this decomposition; actual measurements can also contain fixture leakage and instrument error. [1]

I_{\mathrm{meas}}(t) = I_{\mathrm{charge}}(t) + I_{\mathrm{abs}}(t) + I_{\mathrm{leak}}(t) \tag{1}

Here t is elapsed time after voltage application. Charging and absorption currents may not be separately identifiable from a single terminal-current trace. A specified-time DCL or IR measurement is therefore an operational test result, not proof that every transient has vanished.

Decaying capacitor current with a specified measurement time and corresponding acceptance level — Figure 7. A defined measurement time makes an IR or leakage-current result reproducible.

Use the voltage, temperature, timing and conditioning specified for the particular component. Murata gives 60 s as an example for MLCC insulation-resistance measurement. TDK’s aluminium-electrolytic guidance describes an acceptance measurement after 5 min at 20 °C, while individual series may specify different conditions. There is no single IEC or military measurement time applicable to every capacitor. [1, 2]

Avoiding Measurement Errors

Use a suitable electrometer, picoammeter or source-measure unit, with input protection and current compliance appropriate to the capacitor and its stored energy. Clean, dry fixtures and electrostatic shielding help prevent external leakage and noise from dominating the result. For very high resistance measurements, use guarded connections where supported by the instrument and fixture design. [4, 5]

Check the background current of the fixture at the intended voltage, without the capacitor connected. Confirm that the measurement system remains stable with a capacitive load and that the device reaches the required terminal voltage. Report the result together with voltage, temperature, elapsed time, preconditioning and any background correction.

Enclose hazardous test connections and prevent access while energised. After testing, discharge through an appropriately rated path, verify the terminal voltage and allow for voltage recovery caused by dielectric absorption before handling. Do not rely on a brief short circuit as a complete discharge procedure. [5]

The Time Constant

A charged capacitor with no intentional external connection can lose charge through its internal insulation resistance. If capacitance C and insulation resistance R_IR are approximately constant, the self-discharge time constant τ_leak is given by Eq. (2).

\tau_{\mathrm{leak}} = R_{\mathrm{IR}} C \tag{2}

With initial voltage V₀, the corresponding terminal voltage V_C(t) follows Eq. (3). After one time constant, voltage has fallen to approximately 36.8% of its initial value. This is a first-order model: voltage-dependent leakage, dielectric absorption and external loading can produce a different discharge curve.

V_C(t) = V_0 \exp\left(-\frac{t}{R_{\mathrm{IR}} C}\right) \tag{3}

Exponential capacitor voltage decay illustrating the discharge time constant — Figure 8. Illustration of the first-order discharge time constant.

The product Ω·F has the dimension of time. In capacitor specifications, 1 MΩ·µF equals 1 s; MΩ·µF must not be confused with MΩ·F. Datasheets may specify a minimum resistance for one capacitance range and a minimum CR product for another. Apply the actual limit wording, including any instruction to use the lower of two resistance limits. [1]

Within a family using comparable dielectric material and thickness, increasing electrode area increases capacitance and decreases bulk insulation resistance. This explains the usefulness of a CR-product specification. It is not an exact rule across unrelated dielectrics, voltage ratings or package constructions. Surface leakage and encapsulation can limit the measured resistance, particularly when the intrinsic dielectric resistance is very high. [1]

As a worked example, Murata gives a minimum CR product of 500 Ω·F for a capacitance of 1 µF. Equation (4) converts that requirement into a minimum insulation resistance of 500 MΩ. The result applies at the specification’s stated test conditions; it is not an independent lifetime or charge-retention guarantee. [1]

R_{\mathrm{IR,min}} = \frac{500\ \Omega\cdot\mathrm{F}}{1\times10^{-6}\ \mathrm{F}} = 5\times10^{8}\ \Omega = 500\ \mathrm{M}\Omega \tag{4}

Leakage Current and Its Effects on Capacitor Performance

Leakage matters because it consumes current, changes stored charge and can disturb circuit bias. Its importance depends on the application: a current that is acceptable in a power filter may be unacceptable at a high-impedance sensor input.

In timing, sample-and-hold and high-impedance coupling circuits, leakage can cause voltage drift or DC errors. In battery-powered equipment, it contributes to standby consumption. In power-supply smoothing and decoupling, leakage is one consideration alongside effective capacitance, ESR, ripple-current capability, thermal conditions and useful life. Aluminium and tantalum electrolytic capacitors are not generally unsuitable for power supplies; these are established applications for appropriately selected products. [2, 6]

Film and ceramic capacitors are often useful where low leakage is important, but selection must also address capacitance, bias dependence, absorption, voltage rating and construction. A dielectric label alone does not establish whether a particular component meets a circuit’s leakage budget.

Leakage resistance is a parallel-loss parameter. It must not be confused with ESR, which represents series losses under the relevant AC measurement conditions. Inductance principally follows current-path geometry; these parameters are not all evidence of dielectric defects. [2]

Leakage Current versus Insulation Resistance

At the specified measurement time and operating point, insulation resistance R_IR is the capacitor terminal voltage V_C divided by the measured DCL current I_DCL. Equation (5) gives both forms of this relationship. [1, 4]

R_{\mathrm{IR}} = \frac{V_C}{I_{\mathrm{DCL}}}, \qquad I_{\mathrm{DCL}} = \frac{V_C}{R_{\mathrm{IR}}} \tag{5}

At the same voltage and test conditions, a higher IR corresponds to a lower DCL. Conductance, not current, is the direct reciprocal of resistance. If leakage is nonlinear with voltage, the resistance calculated from Eq. (5) is an operating-point value rather than a voltage-independent material constant.

Original overview comparing insulation resistance across capacitor dielectric types — Figure 9. Original comparative overview of insulation resistance across capacitor dielectric types. Use individual datasheets for selection limits.

Figure 9 provides a broad technology comparison, not interchangeable guaranteed limits. Capacitance, dielectric thickness, voltage, temperature, test duration and construction must be comparable before numerical IR or DCL values are ranked.

Dependence of Leakage Current on Time

The current following voltage application changes as charging, absorption and any technology-specific conditioning processes develop. A datasheet DCL limit at a stated time can differ from the lower current eventually reached during continuous operation. [1, 2]

For an initially uncharged ideal capacitor supplied by a constant voltage V_S through charging resistance R_S, charge Q(t) follows Eq. (6). Here C is constant capacitance, t is elapsed time and exp denotes the exponential function.

Q(t) = C V_S \left[1-\exp\left(-\frac{t}{R_S C}\right)\right] \tag{6}

The corresponding ideal charging current I_charge(t) is given by Eq. (7). These equations neglect leakage and absorption and assume that inductive effects and source-current limiting do not control the response.

I_{\mathrm{charge}}(t) = \frac{V_S}{R_S}\exp\left(-\frac{t}{R_S C}\right) \tag{7}

The charging time constant is τ_charge = R_SC, not I = RC. After one charging time constant, the voltage and charge have completed approximately 63.2% of the change towards their final values. R_S is the charging-path resistance and must not be substituted for the parallel insulation resistance in Eq. (2).

Original capacitor charge and current curves illustrating an exponential transient — Figure 10. Original capacitor charge and current curves illustrating the charging transient.

For aluminium electrolytics, oxide condition, storage history and voltage application influence the later current decay. Operational leakage is the longer-term current reached under the actual voltage and internal temperature. Ceramic capacitors also show time-dependent measured current because absorption does not disappear instantaneously; their measurement timing must therefore be specified. [1, 2]

Self-healing or defect-isolation behaviour differs among wet electrolytics, solid tantalum capacitors and metallised film capacitors. It must not be treated as a universal repair mechanism or as permission to exceed ratings. A recovered leakage reading does not by itself demonstrate that a stressed component remains suitable for service. [2, 6]

Polymer Tantalum Start-up Transients

Some polymer tantalum capacitors can exhibit anomalous charging currents in addition to ordinary displacement, absorption and leakage currents. NASA testing identifies sensitivity to drying, temperature, voltage application and preconditioning. This matters particularly in fast-starting, current-limited and space electronics, where a component may meet a later DCL limit yet still create a start-up problem. [7]

For such applications, evaluate the actual voltage ramp, source impedance, current limit, environmental history and start-up temperature. Do not extrapolate a single low-current leakage test to surge behaviour, or apply a research screening procedure to production hardware without an agreed qualification plan.

Dependence of Leakage Current on Temperature

Longer-term dielectric leakage generally increases with temperature, so insulation resistance decreases. The magnitude and mathematical form depend on dielectric material, electric field and construction; there is no universal temperature multiplier for all capacitor types. [2, 7]

For aluminium electrolytics, use the manufacturer’s guidance at the relevant internal temperature, including self-heating from ripple current. A room-temperature acceptance limit is not automatically the expected leakage during hot operation. Film and ceramic capacitors also require their specified temperature limits rather than a general comparison with electrolytics.

Tantalum and polymer tantalum behaviour additionally depends on cathode system and environmental history. Humidity exposure, drying and thermal processing can affect leakage or start-up transients in different ways. Do not assume that applying rated voltage for a few minutes universally reverses these effects. [6, 7]

Dependence of Leakage Current on Voltage

Leakage normally increases as applied voltage increases, but the relationship need not be linear. For aluminium electrolytics, the increase becomes pronounced above rated voltage. Excess voltage can promote heat generation, gas formation and damage; surge capability applies only under its specified conditions. [2]

Reducing voltage can reduce leakage, but the reduction must be taken from the relevant product data or measurement. Similarly, ceramic insulation resistance cannot be assumed independent of voltage when conduction is field-dependent. An IR value is meaningful only with its associated test voltage and timing. [1]

After prolonged unpowered storage, some aluminium electrolytics draw increased initial leakage as the oxide condition changes. Reforming is a controlled, manufacturer-defined conditioning process, not a universal half-hour application of rated voltage. Follow the applicable procedure for current limitation, duration, temperature and subsequent measurements; reforming does not restore lost electrolyte or repair mechanical damage. [2]

DCL Myths and Reliability

Leakage provides useful information, but its interpretation depends on the technology and test conditions. Two common simplifications confuse normal conduction with defects and a single electrical measurement with overall reliability.

Myth 1: All Leakage Comes from Cracks in the Dielectric

Cracks, contamination and other defects can increase leakage and cause failure. However, finite leakage also occurs in otherwise serviceable capacitors because real dielectrics and their electrode interfaces support charge transport. A measured current is not a direct count of cracks. [1, 6, 7]

Approximately ohmic conduction gives a current roughly proportional to voltage over the applicable range.
Poole–Frenkel emission describes electric-field-assisted release of trapped carriers in the dielectric. It is not simply another name for every form of hopping conduction.
Schottky emission describes carrier injection influenced by an electrode or cathode interface barrier. Interface properties can therefore change leakage without a corresponding change in the number of dielectric cracks.
Tunnelling is quantum-mechanical transport through a barrier. It must not be equated with avalanche multiplication or described as the universal capacitor breakdown mechanism.

The dominant mechanism depends on dielectric, interfaces, field and temperature. Metal-insulator-metal models are useful simplifications, but oxide capacitors with solid cathodes can require explicit consideration of semiconductor-like barriers and interfacial states. These distinctions explain why a change of cathode material can alter DCL even when the nominal dielectric chemistry remains unchanged. [6, 7]

Myth 2: Leakage Alone Establishes Component Reliability

A lower leakage value does not automatically make one capacitor technology more reliable than another. Reliability depends on construction, manufacturing quality, voltage and thermal stress, source impedance, environment and the relevant failure mechanisms. Conversely, excessive or unstable leakage in a controlled test must not be dismissed as harmless. [2, 6, 7]

Changes in leakage after electrical, thermal or mechanical stress can reveal degradation. Screening and lot acceptance are useful only when the test conditions, limits and demonstrated failure mechanisms support them. A single room-temperature DCL result cannot replace qualification, endurance data or application-level assessment. [7]

Case Study: MnO₂ and Polymer Tantalum Capacitors

Replacing an MnO₂ cathode with a conductive polymer changes the cathode-dielectric interface as well as series resistance and defect-isolation behaviour. A historical JPL study reported polymer-tantalum catalogue leakage limits commonly about ten times those of MnO₂ types. That was a comparison of the products and limits studied, not a universal present-day ratio for measured leakage. [6]

The example remains useful because it shows why greater DCL does not imply proportionally worse reliability or proportionally more cracks. For current selection, compare the actual product limits, temperature and voltage conditions, qualification evidence and start-up behaviour. Polymer-specific transient effects are a separate consideration from the later specified leakage reading. [6, 7]

Design and Selection Checklist

Begin with the circuit’s allowable leakage and operating conditions, then translate them into a component specification and reproducible acceptance test. The same nominal capacitance and voltage rating do not make two capacitor series interchangeable.

Define the application: precision signal storage, high-impedance coupling, timing, battery standby, power decoupling or bulk energy storage. Include leakage from connected semiconductors, the board and any discharge network.
Compare guaranteed maximum DCL or minimum IR, not only typical values. Record test voltage, temperature, electrification time, preconditioning and whether the limit is a resistance or a CR product.
Check effective capacitance, tolerance, dielectric absorption, polarity and voltage dependence where these affect stored charge or signal accuracy.
For power and SMPS applications, also compare ESR, ripple-current capability, thermal conditions, transients and useful-life requirements. Do not use low DCL as a substitute for these checks.
For catalogue parts, use the exact series and ordering-code specification. For custom parts or special screening, agree leakage limits, test timing, conditioning, traceability and change control in the procurement specification.
Match qualification and acceptance testing to the application. Obtain the relevant test evidence rather than inferring reliability from a single leakage value or a qualification label.
Provide safe current limiting, test access protection and controlled discharge. Where safety insulation or mains connection is involved, verify the required approvals independently of terminal-to-terminal IR.

TDK explicitly distinguishes series specifications, custom requirements and qualification test conditions in its technical guidance. For specialist polymer-tantalum applications, the qualification plan may also need to address start-up transients and environmental preconditioning. [2, 7]

Conclusion

Insulation resistance and DC leakage current describe a capacitor’s DC insulation behaviour under specified conditions. Their relationship is straightforward, but interpreting a measurement requires separation of charging and absorption effects, control of voltage and temperature, and attention to fixture leakage and test history.

Select capacitors against the application’s leakage budget and the complete datasheet, not a universal technology ranking. Use leakage trends and acceptance limits as part of reliability assessment, while keeping surge capability, endurance, environmental effects and safe discharge requirements separate.

FAQ

What is insulation resistance in a capacitor?

Insulation resistance is the resistance calculated from a capacitor’s terminal voltage divided by its measured DC current under specified conditions. A higher value means lower current at the same voltage. Ceramic and film capacitors commonly specify a minimum resistance or a minimum capacitance-resistance product. Test voltage, temperature and measurement time must accompany the value.

What is DC leakage current and how does it relate to IR?

DC leakage current is the current remaining under an applied DC voltage after charging and absorption effects have sufficiently diminished for the specified test. Insulation resistance equals terminal voltage divided by that current. Current and resistance are therefore not direct reciprocals. Comparisons require the same voltage, temperature, timing and conditioning, especially when conduction is nonlinear.

How long should I wait before measuring capacitor leakage?

Use the electrification time specified in the capacitor datasheet or applicable acceptance procedure. There is no universal waiting time for every technology. The initial current includes charging and absorption components, so an early reading may overstate longer-term leakage. Record the elapsed time and actual terminal voltage, and apply any required preconditioning before comparing the result with a limit.

What factors affect capacitor leakage current?

Leakage depends on dielectric material, construction, applied voltage, temperature and operating history. The measured current also depends on the time after voltage application because charging and absorption currents decay. Storage, humidity and thermal processing can matter for particular technologies. Fixture contamination, board leakage and instrument background can additionally affect the result without representing internal capacitor leakage.

Why is leakage current important when selecting a capacitor?

Leakage consumes standby current, reduces retained charge and can create errors in high-impedance signal or timing circuits. Its importance depends on the circuit’s current budget and required accuracy. For power filtering, leakage remains relevant but must be assessed alongside capacitance, ESR, ripple-current capability and thermal life. Electrolytic capacitors are not generally excluded from power-supply applications.

Does high leakage current mean a capacitor is unreliable?

An out-of-limit or unstable leakage result requires investigation, but leakage alone does not rank reliability across different capacitor technologies. Construction and cathode interfaces can produce different normal leakage levels. Changes after controlled stress can reveal degradation. Reliability assessment also needs relevant qualification, endurance and application data, with measurement conditions controlled before conclusions are drawn.

Can reforming restore an old aluminium electrolytic capacitor?

Controlled reforming can reduce elevated leakage associated with oxide condition after storage, when the manufacturer permits the procedure. It is not a general repair method and does not restore lost electrolyte or mechanical damage. Follow the specified current limitation, voltage, duration and subsequent tests. Do not apply a universal rated-voltage treatment to an unknown or visibly damaged component.

How to Measure Capacitor Insulation Resistance and Leakage Current

Measure capacitor leakage current and calculate insulation resistance under the component’s specified test conditions using a protected, current-limited setup.

Define the test conditions
Identify the exact component and obtain its specified test voltage, temperature, electrification time, leakage or IR limit and preconditioning requirements. Do not assume that rated voltage is always the required test voltage.
Prepare a protected measurement setup
With the source off, verify that the capacitor is discharged. Connect a suitable current-measurement instrument and current-limited source using the correct polarity. Use clean fixtures, shielding and guarding as appropriate, and prevent access to hazardous connections.
Check background and apply the test voltage
Check fixture background current at the intended voltage before connecting the capacitor. Apply the specified voltage through the protected setup, start timing and confirm the actual capacitor terminal voltage. Stop if current, heating or other behaviour exceeds the permitted conditions.
Record current at the specified time
Allow the required electrification interval and record current, terminal voltage and temperature. Distinguish the initial charging response from the later reading. If required, retain the current-time trace to identify incomplete settling, current limiting or abnormal transients.
Calculate IR and compare the result
Divide the measured capacitor terminal voltage by the measured current to calculate insulation resistance. Compare DCL or IR with the applicable datasheet limit under matching conditions. Include any specified CR-product conversion and document background correction and measurement limitations.
Discharge and verify safe handling
Remove the applied voltage and discharge through a suitably rated path. Verify the terminal voltage before disconnecting or handling the capacitor, and allow for voltage recovery caused by dielectric absorption. Record the test conditions and result for traceability.

References

The following manufacturer guidance and technical reports support the definitions, measurement practices and technology-specific discussion. Historical reports are used for the stated historical comparisons, not as current product specifications.

Murata Manufacturing, Insulation Resistance and Leakage Current of Ceramic Capacitor, Murata Capacitor Guide, 2023.
TDK Electronics, Aluminum Electrolytic Capacitors: General Technical Information, TDK Electronics AG, August 2026.
Peter Matthews, Fundamental Capacitor Functions in the World of Electronics, Knowles Precision Devices Blog, 2024.
Keithley Instruments, Leakage Current Test and Insulation Resistance Measurements, Tektronix, undated web edition.
Keysight Technologies, Capacitance Leakage Current Measurement Techniques Using the B2985A/87A, Keysight Technologies, undated web edition.
Erik K. Reed, Characterization of Tantalum Polymer Capacitors, Jet Propulsion Laboratory, NASA Electronic Parts and Packaging Program, FY2005 report.
Alexander A. Teverovsky, Test Methods and Analysis of Anomalous Charging Currents in Polymer Tantalum Capacitors, NASA/TM—20220010090, NASA Goddard Space Flight Center, October 2022.

Latest technical review and revision: 27 October 2026. Definitions, equations, measurement guidance, application scope and reliability discussion updated.