Energy Conditions

Einstein's equation $G_{\mu\nu}=8\pi T_{\mu\nu}$ is a two-way street with a fatal loophole. Read left to right, matter tells spacetime how to curve. Read right to left, it lets you manufacture any spacetime you like — pick a metric, compute its Einstein tensor, and define the matter to be whatever $T_{\mu\nu}=G_{\mu\nu}/8\pi$ comes out. Wormholes, time machines, warp bubbles are all exact solutions of the field equation. What rules them out is not the equation but a judgement about which stress-energy tensors count as physically reasonable matter. Energy conditions are that judgement, written as coordinate-invariant inequalities on $T_{\mu\nu}$ . They form a graded family, distinguished by which vectors are contracted against $T_{\mu\nu}$ and whether they constrain energy density, energy flux, or the gravitating combination that includes pressure. The null (NEC), weak (WEC), dominant (DEC), and strong (SEC) conditions each answer a physically distinct question — is energy density non-negative along a light ray, for every observer, does energy flow slower than light, does gravity stay attractive — and each is the precise hypothesis some global theorem requires. Their power is that, through the Einstein equation, an inequality on matter becomes an inequality on the Ricci tensor, and a Ricci inequality of definite sign is exactly what the Raychaudhuri equation needs to force a bundle of geodesics to focus — the engine behind the singularity theorems, the area law, the positive-mass theorem, and topological censorship. The twist is that energy conditions are not laws of nature. Classical matter obeys them; quantum fields routinely do not — the Casimir vacuum has negative energy density, Hawking radiation lets black holes shrink, and a positive cosmological constant drives the universe apart while violating the strong condition. This tension — pointwise conditions that fail, yet theorems that must survive — has reshaped the modern subject into averaged and quantum energy inequalities that bound how negative energy can get and for how long.

Key takeaways

Energy conditions are pointwise inequalities on the stress-energy tensor $T_{\mu\nu}$ encoding that energy densities are non-negative and that energy does not flow faster than light; the hierarchy — null (NEC), weak (WEC), dominant (DEC), strong (SEC) — is set by which observers and contractions are constrained.
Through $G_{\mu\nu}=8\pi T_{\mu\nu}$ , a condition on $T_{\mu\nu}$ becomes a sign condition on $R_{\mu\nu}$ — the null convergence condition from NEC, the timelike one from SEC — which is the source term in the Raychaudhuri equation.
Definite-sign Ricci curvature forces geodesic congruences to focus to a caustic in finite affine parameter: the mechanism behind the Penrose–Hawking singularity theorems, the area theorem, and the positive-mass theorem.
DEC $\Rightarrow$ WEC $\Rightarrow$ NEC and SEC $\Rightarrow$ NEC, but SEC and WEC are logically independent — the "strong" condition does not imply the "weak" one.
Quantum fields violate all of them pointwise (Casimir energy is negative; $\Lambda>0$ breaks SEC), rescued in modern work by averaged conditions like ANEC and by quantum energy inequalities.
Key idea: Each energy condition is a sign requirement on a particular contraction of $T_{\mu\nu}$ with timelike or null vectors; through $G_{\mu\nu}=8\pi T_{\mu\nu}$ it becomes a sign requirement on $R_{\mu\nu}$ , which is exactly the source term in the Raychaudhuri equation that forces geodesics to focus.

Where this lives today

Energy conditions have moved from the chalkboard to the data. In 2021 Isi, Farr, and collaborators used the ringdown of GW150914 to test Hawking's area theorem directly, confirming at better than 95% confidence that the final horizon area exceeds the sum of the initial areas — an observational check of a theorem whose proof rests on the null energy condition. On the theory side, the pointwise conditions have been superseded by sharper quantum statements: quantum energy inequalities bound the magnitude times duration of any negative-energy excursion, the averaged null energy condition has been proven from the unitarity and causality of quantum field theory, and the quantum null energy condition (QNEC), relating energy density to the second derivative of entanglement entropy, was conjectured by Bousso and then proven. These matter because the frontier questions all hinge on how much a condition can be violated: whether a traversable wormhole can be engineered (only microscopically, as in the Gao–Jafferis–Wall construction), whether the Alcubierre warp drive can be built (its negative-energy budget is astronomically large), and whether the Big Bang singularity is real or an artefact of assuming the strong condition dark energy already violates. The humble inequality $T_{\mu\nu}k^\mu k^\nu\ge0$ sits at the exact boundary between what classical gravity permits and what quantum matter allows.

The physical picture

Picture a small ball of freely-falling test particles, each on its own geodesic, released at rest. Ordinary matter attracts, so the ball can only shrink — its volume never spontaneously grows. The energy conditions are the exact statement of "gravity is attractive": they fix the sign of the Ricci term in the Raychaudhuri equation so the bundle of worldlines converges rather than diverges. Follow the convergence far enough and neighbouring geodesics cross at a caustic, a focal point where the smooth congruence breaks down — the local seed of a singularity, and what the global theorems are built to force into existence.

Figure

A congruence of timelike geodesics under attractive gravity. When the Ricci contraction Rμν uᵘuᵛ ≥ 0 the expansion θ can only decrease, so an initially converging bundle with negative θ₀ shrinks in cross-section and focuses to a caustic in finite affine parameter — the local mechanism behind geodesic incompleteness

The four conditions are four answers to "is the source attractive", each contracting $T_{\mu\nu}$ against a different class of vectors; they are partly nested, but the strong condition can fail while the others hold. The cleanest way to see the whole family is to reduce each to an inequality on the density $\rho$ and pressure $p$ of a perfect fluid and plot the allowed regions.

Figure

The energy-condition hierarchy in the density–pressure plane for a perfect fluid. NEC needs ρ+p ≥ 0, WEC adds ρ ≥ 0, DEC needs −ρ ≤ p ≤ ρ, SEC needs ρ+3p ≥ 0. Dark energy with p = −ρ lies on the NEC/DEC edge yet falls below the SEC boundary p = −ρ⁄3, so ρ+3p = −2ρ is negative and the strong condition alone fails

Why does pressure enter the strong condition with triple weight? Because in relativity pressure gravitates: in Newton's theory only mass density sources the field, but in Einstein's the source for timelike focusing is $\rho+3p$ . A photon gas ( $p=\rho/3$ ) gravitates harder than its energy density alone would suggest, while a medium under large tension — negative pressure, as for a cosmological constant or inflaton — makes $\rho+3p$ negative, flips the sign of the Ricci source, and produces repulsion. That single sign flip is why cosmic acceleration is, technically, a violation of the strong energy condition.

The core equations

The null energy condition is the weakest and most robust member: it contracts $T_{\mu\nu}$ against null vectors only, so it constrains the least and survives where the others fail. It is the hypothesis of the area theorem and of most wormhole no-go results.

T_{\mu\nu}k^\mu k^\nu \ge 0 \quad \text{for all null } k^\mu \qquad (1)

where: $T_{\mu\nu}$ — stress-energy tensor, symmetric, carrying energy density and flux; $k^\mu$ — arbitrary future-directed null vector, $g_{\mu\nu}k^\mu k^\nu=0$ . Reading: along any light ray the effective energy density cannot be negative — for a perfect fluid, just $\rho+p\ge0$ . The weak energy condition strengthens this to all timelike observers: contracting against a unit four-velocity gives the density that observer measures, so WEC says "no observer ever sees negative energy density".

T_{\mu\nu}t^\mu t^\nu \ge 0 \quad \text{for all timelike } t^\mu \qquad (2)

where: $t^\mu$ — arbitrary future-directed timelike vector, $g_{\mu\nu}t^\mu t^\nu<0$ in signature $-+++$ . Reading: normalizing $t^\mu t_\mu=-1$ , the contraction $\rho_{\rm obs}=T_{\mu\nu}t^\mu t^\nu$ is the measured density, and WEC demands it be non-negative for every observer — for a perfect fluid, $\rho\ge0$ and $\rho+p\ge0$ . The null limit $t^\mu\to k^\mu$ recovers NEC, so WEC $\Rightarrow$ NEC. The dominant energy condition adds a causality clause: energy density is non-negative and the measured energy-momentum flux is future-directed causal, so energy never streams faster than light.

T_{\mu\nu}t^\mu t^\nu \ge 0 \quad\text{and}\quad J^\mu=-T^\mu{}_\nu t^\nu \ \text{is future-directed causal} \qquad (3)

where: $J^\mu$ — the energy-momentum flux four-vector observer $t^\mu$ measures; "causal" means $J^\mu J_\mu\le0$ . Reading: for a perfect fluid this collapses to $\rho\ge|p|$ . DEC is the central hypothesis of the positive-mass theorem, and DEC $\Rightarrow$ WEC $\Rightarrow$ NEC. The strong energy condition looks different: it constrains the trace-reversed stress-energy tensor, not $T_{\mu\nu}$ itself — which makes it a statement about gravitational attraction rather than energy density, and is why the name is a trap: SEC does not imply WEC.

\left(T_{\mu\nu}-\tfrac{1}{2}g_{\mu\nu}T\right)t^\mu t^\nu \ge 0 \quad \text{for all timelike } t^\mu \qquad (4)

where: $T=g^{\mu\nu}T_{\mu\nu}=-\rho+3p$ for a perfect fluid; the combination $T_{\mu\nu}-\tfrac12 g_{\mu\nu}T$ equals $R_{\mu\nu}/8\pi$ by the field equation. Reading: SEC reduces to $\rho+p\ge0$ and $\rho+3p\ge0$ , the $\rho+3p$ being the fingerprint of pressure's triple gravitational weight. The bridge that gives all this its teeth is the Einstein equation, which converts every matter inequality into a curvature inequality — immediately for null vectors, because the metric term drops out.

G_{\mu\nu}=8\pi T_{\mu\nu}\;\Longrightarrow\; R_{\mu\nu}k^\mu k^\nu=8\pi\,T_{\mu\nu}k^\mu k^\nu,\qquad R_{\mu\nu}t^\mu t^\nu=8\pi\!\left(T_{\mu\nu}-\tfrac12 g_{\mu\nu}T\right)t^\mu t^\nu \qquad (5)

where: $G_{\mu\nu}=R_{\mu\nu}-\tfrac12 g_{\mu\nu}R$ — Einstein tensor; $R_{\mu\nu}$ — Ricci tensor. Reading: NEC becomes the null convergence condition $R_{\mu\nu}k^\mu k^\nu\ge0$ with no trace reversal (since $g_{\mu\nu}k^\mu k^\nu=0$ ), while SEC becomes $R_{\mu\nu}t^\mu t^\nu\ge0$ — precisely the sign conditions the focusing equation demands. That focusing equation is the Raychaudhuri equation, the exact evolution law for a congruence's expansion; its Ricci term is where the energy conditions enter the dynamics.

\frac{\mathrm{d}\theta}{\mathrm{d}\lambda}=-\frac{1}{n-1}\theta^2-\sigma_{\mu\nu}\sigma^{\mu\nu}+\omega_{\mu\nu}\omega^{\mu\nu}-R_{\mu\nu}u^\mu u^\nu \qquad (6)

where: $\theta$ — expansion (fractional rate of change of cross-sectional volume); $\sigma_{\mu\nu}$ — shear; $\omega_{\mu\nu}$ — vorticity; $\lambda$ — affine parameter; $u^\mu$ — tangent; $n-1$ — transverse dimension, 3 for timelike and 2 for null congruences in 4D. Reading: shear and the convergence-condition Ricci term both decrease $\theta$ (focusing), while vorticity increases it (rotation resists collapse) — so the theorems require vorticity-free congruences. Quantum fields break the pointwise conditions; the canonical demonstration is the Casimir vacuum, where the renormalized energy density between two conducting plates is negative.

\langle 0|T_{00}|0\rangle_{\rm Casimir}=-\frac{\pi^2}{720}\frac{\hbar c}{a^4},\qquad \frac{F}{A}=-\frac{\pi^2}{240}\frac{\hbar c}{a^4} \qquad (7)

where: $a$ — plate separation; $\hbar,c$ — reduced Planck constant and light speed, restored from geometrized units. Reading: a comoving observer measures $\langle T_{00}\rangle<0$ , a direct pointwise violation of WEC and NEC; the companion force per area is the measured Casimir force. The modern repair is to average along a complete geodesic: the averaged null energy condition asks only that the integral be non-negative, which can hold even when the pointwise version fails everywhere.

\int_{-\infty}^{\infty} T_{\mu\nu}k^\mu k^\nu\,\mathrm{d}\lambda \ge 0 \qquad (8)

where: $\lambda$ — affine parameter along a complete null geodesic; $k^\mu=\mathrm{d}x^\mu/\mathrm{d}\lambda$ . Reading: ANEC, proven for free and many interacting quantum field theories, powers topological censorship and singularity-type results without any pointwise assumption — which is why it is the physically preferred statement.

Derivation AFrom an energy condition to geodesic focusing

The goal is to turn "attractive gravity" into a hard theorem: if the convergence condition $R_{\mu\nu}u^\mu u^\nu\ge0$ holds and the congruence is irrotational, any initially converging bundle of timelike geodesics reaches a caustic ( $\theta\to-\infty$ ) within a finite, explicitly bounded amount of affine parameter. This is the central consequence of the energy conditions, because a caustic where geodesics cross is the local obstruction the Penrose–Hawking theorems parlay into geodesic incompleteness. We start from the full Raychaudhuri equation and strip it down.

Step 1. For a hypersurface-orthogonal congruence the Frobenius theorem gives $\omega_{\mu\nu}=0$ , and in 4D the transverse dimension is $n-1=3$ , so equation (6) becomes $\mathrm{d}\theta/\mathrm{d}\lambda=-\tfrac13\theta^2-\sigma_{\mu\nu}\sigma^{\mu\nu}-R_{\mu\nu}u^\mu u^\nu$ . Why: vorticity is the only term that could fight collapse, so removing it leaves every remaining term with a definite sign.

Step 2. Bound the right-hand side. A spacelike shear has $\sigma_{\mu\nu}\sigma^{\mu\nu}\ge0$ and the convergence condition gives $R_{\mu\nu}u^\mu u^\nu\ge0$ , so both subtracted terms are non-positive and $\mathrm{d}\theta/\mathrm{d}\lambda\le-\tfrac13\theta^2$ . Why: discarding shear and Ricci as help keeps only the inescapable $-\tfrac13\theta^2$ ; true focusing is at least this fast.

Step 3. Integrate. Dividing by $\theta^2$ gives $\mathrm{d}(\theta^{-1})/\mathrm{d}\lambda\ge\tfrac13$ , so $\theta^{-1}(\lambda)\ge\theta_0^{-1}+\lambda/3$ . Why: the nonlinear inequality for $\theta$ is linear in $\theta^{-1}$ , which integrates trivially and exposes the finite-time blow-up.

Step 4. Read off the caustic. If the congruence starts converging, $\theta_0<0$ , then $\theta^{-1}$ starts negative and climbs to zero, so $\theta\to-\infty$ at or before $\lambda=3/|\theta_0|$ — a caustic where geodesics cross and the cross-section collapses. Why: attractive curvature plus initial convergence guarantees focusing in bounded affine parameter, whatever the detailed matter does. *Note: the transverse factor 3 for timelike congruences is replaced by 2 for null ones, giving the tighter bound $\Delta\lambda\le2/|\theta_0|$ used in the area theorem.*

Go deeper · Derivation B · 1 minThe strong energy condition for a perfect fluid

Derivation BThe strong energy condition for a perfect fluid

The goal is to compute the trace-reversed contraction of equation (4) explicitly for a perfect fluid, extract the pair $\rho+p\ge0$ and $\rho+3p\ge0$ , and show how pressure acquires its triple gravitational weight and why a positive cosmological constant violates the very condition it most directly controls. We use $T^{\mu\nu}=(\rho+p)u^\mu u^\nu+p\,g^{\mu\nu}$ and a comoving observer $u^\mu u_\mu=-1$ , then generalize.

Step 1. Take the trace: $T=g^{\mu\nu}T_{\mu\nu}=(\rho+p)(-1)+p\cdot4=-\rho+3p$ , using $g^{\mu\nu}g_{\mu\nu}=4$ and $u^\mu u_\mu=-1$ . Why: the trace is the one new ingredient SEC needs, and it is where the pressure's factor of 3 is born — four dimensions minus the timelike direction.

Step 2. Contract the raw stress-energy: $T_{\mu\nu}u^\mu u^\nu=(\rho+p)(-1)^2+p(-1)=\rho$ . Why: the comoving observer simply measures $\rho$ , so any correction to SEC comes entirely from the trace-reversal term.

Step 3. Form the metric piece: $-\tfrac12 g_{\mu\nu}u^\mu u^\nu\,T=-\tfrac12(-1)(-\rho+3p)=\tfrac12(-\rho+3p)$ . Why: for a timelike vector this metric term is nonzero — precisely the term that vanishes for null vectors and so distinguishes SEC from NEC.

Step 4. Add Steps 2 and 3: $\left(T_{\mu\nu}-\tfrac12 g_{\mu\nu}T\right)u^\mu u^\nu=\rho+\tfrac12(-\rho+3p)=\tfrac12(\rho+3p)\ge0$ , giving $\rho+3p\ge0$ ; a boosted observer plus the null limit yields the companion $\rho+p\ge0$ . Why: together these are SEC, and by equation (5) they are exactly $R_{\mu\nu}u^\mu u^\nu\ge0$ , the timelike convergence condition the focusing theorem consumes. *A positive cosmological constant has $\rho_\Lambda=\Lambda/8\pi>0$ and $p_\Lambda=-\rho_\Lambda$ , so $\rho+3p=-2\rho_\Lambda<0$ and SEC fails — confirming that $\Lambda>0$ produces gravitational repulsion — while $\rho+p=0$ keeps NEC, WEC, and DEC intact.*

Go deeper · Derivation C · 1 minThe logical hierarchy of the four conditions

Derivation CThe logical hierarchy of the four conditions

The goal is to prove the chain DEC $\Rightarrow$ WEC $\Rightarrow$ NEC and SEC $\Rightarrow$ NEC, and to show SEC and WEC are genuinely independent, so neither the naming order nor any naive intuition about "strength" survives scrutiny. We work in the rest frame of a type-I (non-null) stress-energy, where symmetry lets us diagonalize it to principal values $(\rho,p_1,p_2,p_3)$ and read each condition off directly.

Step 1. WEC in the eigenframe. Demanding $T_{\mu\nu}t^\mu t^\nu\ge0$ for every timelike $t^\mu$ reduces, after maximizing over boosts, to $\rho\ge0$ and $\rho+p_i\ge0$ for each $i$ . Why: these are the eigenframe statements of "every observer sees non-negative density", the reference for the other conditions.

Step 2. WEC $\Rightarrow$ NEC. Sending $t^\mu$ to a null direction suppresses the isolated $\rho$ term and leaves $\rho+p_i\ge0$ , exactly NEC. Why: NEC is the null-limit boundary case of WEC, so any WEC tensor obeys NEC — the first link of the chain.

Step 3. DEC $\Rightarrow$ WEC. The causal-flux clause $J^\mu J_\mu\le0$ sharpens the constraint to $\rho\ge|p_i|$ , which implies both $\rho\ge0$ and $\rho+p_i\ge0$ . Why: forbidding superluminal energy flow is strictly stronger than non-negative density, so DEC tops the density/flux hierarchy.

Step 4. SEC $\Rightarrow$ NEC but not WEC. SEC in the eigenframe reads $\rho+p_i\ge0$ (which is NEC) plus $\rho+\sum_j p_j\ge0$ , but never requires $\rho\ge0$ alone. Why: a medium with $\rho<0$ and large pressures can satisfy SEC while flunking WEC, while $\Lambda>0$ does the reverse — the two are logically independent, and "strong" versus "weak" refers to historical strength of consequence, not implication. *The independence of SEC and WEC is exhibited concretely by $\Lambda>0$ on one side — which satisfies WEC but violates SEC — and a hypothetical medium with $\rho<0,\,p>|\rho|$ on the other, which can satisfy SEC while violating WEC.*

Limiting cases

Each condition is tested here on a perfect fluid $T^{\mu\nu}=(\rho+p)u^\mu u^\nu+p\,g^{\mu\nu}$ , with $\rho$ the rest-frame energy density and $p$ the isotropic pressure:

Case	Result	Meaning
Null (NEC)	$\rho+p\ge0$	Weakest condition; the density-plus-pressure combination that drives null focusing and underlies the area theorem.
Weak (WEC)	$\rho\ge0$ and $\rho+p\ge0$	Every observer measures non-negative energy density. WEC $\Rightarrow$ NEC. Violated by quantum vacua such as the Casimir effect.
Dominant (DEC)	$\rho\ge0$ and $-\rho\le p\le\rho$	Energy-momentum flux is future causal — no superluminal energy flow. DEC $\Rightarrow$ WEC $\Rightarrow$ NEC. Hypothesis of the positive-mass theorem.
Strong (SEC)	$\rho+p\ge0$ and $\rho+3p\ge0$	Controls $R_{\mu\nu}t^\mu t^\nu$ ; violated by $\Lambda>0$ ( $\rho+3p=-2\rho<0$ ) and by inflation. Implies NEC but NOT WEC.
Cosmological constant ( $p=-\rho$ )	NEC, WEC, DEC hold; SEC fails	Saturates NEC and DEC; drives accelerated expansion precisely because $\rho+3p<0$ .
Phantom energy ( $p<-\rho$ )	all four fail	$\rho+p<0$ breaks even NEC; associated with Big-Rip futures and genuinely exotic matter.

Worked examples

1 · Classifying cosmological fluids (confidence builder)

Model cosmic matter as a perfect fluid with equation of state $p=w\rho$ , $\rho>0$ . For dust ( $w=0$ ), radiation ( $w=1/3$ ), a cosmological constant ( $w=-1$ ), and phantom energy ( $w=-4/3$ ), determine which of NEC, WEC, DEC, SEC each satisfies.

Step 1. In terms of $w$ : NEC is $(1+w)\rho\ge0$ , i.e. $w\ge-1$ ; WEC adds $\rho\ge0$ so it too needs $w\ge-1$ ; DEC is $|p|\le\rho$ , i.e. $-1\le w\le1$ ; SEC is $(1+3w)\rho\ge0$ , i.e. $w\ge-1/3$ .

Step 2. Evaluate. Dust $w=0$ and radiation $w=1/3$ satisfy all four. Cosmological constant $w=-1$ : NEC/WEC/DEC saturated, SEC fails since $-1<-1/3$ . Phantom $w=-4/3<-1$ : all four fail, starting with NEC.

Step 3. Interpret. The number that flips is $1+3w$ : positive for ordinary matter and radiation (attractive gravity, deceleration), negative for dark energy (repulsion, acceleration). Only the strong condition distinguishes accelerating from decelerating cosmologies; NEC is the last line of defence, crossed only by phantom energy.

2 · Negative energy in the Casimir vacuum (fully worked)

Two parallel conducting plates sit $a=1\ \mu\mathrm{m}$ apart in vacuum. Compute the vacuum energy density between them and the pressure, and say which energy conditions are violated.

Step 1. In equation (7), $\hbar c=1.055\times10^{-34}\times2.998\times10^8=3.16\times10^{-26}\ \mathrm{J\,m}$ and $a^4=10^{-24}\ \mathrm{m}^4$ , so $\hbar c/a^4=3.16\times10^{-2}\ \mathrm{J/m^3}$ ; the coefficient is $\pi^2/720=1.371\times10^{-2}$ .

Step 2. Energy density: $\langle T_{00}\rangle=-1.371\times10^{-2}\times3.16\times10^{-2}=\mathbf{-4.3\times10^{-4}\ J/m^3}$ — negative: a comoving observer between the plates measures a below-vacuum energy density.

Step 3. Pressure: with $\pi^2/240=4.112\times10^{-2}$ , $F/A=-4.112\times10^{-2}\times3.16\times10^{-2}=\mathbf{-1.3\ mPa}$ , an attraction pulling the plates together — the quantity Lamoreaux measured to a few percent in 1997.

Step 4. Verdict. Since $\langle T_{00}\rangle<0$ , the comoving observer sees $T_{\mu\nu}t^\mu t^\nu<0$ : WEC is violated, and in the null limit NEC too. What survives is the average — integrated along a complete null geodesic the negative Casimir region is offset by the positive-energy plates, so ANEC still holds, which is why the averaged singularity theorems are not destroyed by quantum fields. Between 1 μm plates the vacuum energy density is −4.3×10⁻⁴ J/m³ and the pressure −1.3 mPa: WEC and NEC fail pointwise, ANEC survives.

3 · Free-fall collapse from the Raychaudhuri equation (fully worked)

A ball of pressureless dust at the Sun's mean density $\rho=1.41\times10^3\ \mathrm{kg/m^3}$ is released from rest. Estimate the time to total collapse, and identify which energy condition guarantees it.

Step 1. Because $p=0$ , SEC reads $\rho+3p=\rho>0$ , so $R_{\mu\nu}u^\mu u^\nu=4\pi G\rho>0$ (restoring units) and the comoving congruence must focus. Released from rest, $\theta_0=0$ , but $\mathrm{d}\theta/\mathrm{d}\tau=-4\pi G\rho<0$ immediately drives $\theta$ negative.

Step 2. The collapse this focusing describes has the free-fall time $t_{\rm ff}=\sqrt{3\pi/(32 G\rho)}$ . Insert $G=6.674\times10^{-11}$ and $\rho=1.41\times10^3$ : $32G\rho=3.01\times10^{-6}$ , so $3\pi/(32G\rho)=3.13\times10^{6}\ \mathrm{s^2}$ .

Step 3. Take the root: $t_{\rm ff}=\mathbf{1.77\times10^{3}\ s\approx30\ min}$ . The caustic where $\theta\to-\infty$ is the collapse of the ball to a point.

Step 4. Interpret. The energy condition did the essential work — it fixed the sign of the Ricci source so Raychaudhuri could only converge, and the finite-time caustic is the same phenomenon that, given a trapped surface, becomes the Penrose singularity theorem. The strong condition forces a solar-density dust ball to collapse in about 30 minutes — the free-fall time recovered directly from geodesic focusing.

Physics of the portrait

Energy conditions are inseparable from the person who first showed what they could do. For fifty years the open question was whether the singularities in the known exact solutions were mere artefacts of symmetry — and one man answered it, by inventing the machinery that turns an energy condition into a theorem.

Origins · 1 minRoger Penrose and how this was discovered

Roger Penrose

b. 1931, English mathematical physicist · singularity theorems and global methods in general relativity

Roger Penrose changed general relativity from a subject about special solutions into one about global structure. Before his 1965 paper, the singularities in the Schwarzschild and Friedmann metrics were widely dismissed as artefacts of exact spherical symmetry — remove the symmetry, the reasoning went, and infalling matter would swirl, miss the centre, and bounce. Penrose dissolved this hope with a single new idea, the closed trapped surface: a two-surface so deep in a gravitational field that both its outgoing and ingoing light fronts converge. Combining such a surface with the null energy condition (via Raychaudhuri focusing of the null generators) and a non-compact Cauchy surface, he proved the spacetime must be null-geodesically incomplete — a singularity forms, symmetry or no symmetry. The proof used almost nothing about the matter beyond one inequality, $T_{\mu\nu}k^\mu k^\nu\ge0$ , which is exactly why energy conditions became foundational rather than incidental. With Stephen Hawking he extended it into the Hawking–Penrose theorem covering collapse and cosmology alike, showing that classical general relativity generically predicts its own breakdown. The 2020 Nobel Prize recognized the 1965 theorem as "the discovery that black-hole formation is a robust prediction of the general theory of relativity", more than half a century after the eight-page paper that started it.

History

1948

— Casimir predicts a negative vacuum energy density between conducting plates

1955

— Raychaudhuri derives the focusing equation for geodesic congruences

1965

— Penrose proves the first singularity theorem from the null convergence condition

1971

— Hawking proves the black-hole area theorem using the null energy condition

1973

— Hawking and Ellis systematize the energy-condition hierarchy in their monograph

Going further

Texts. Misner, Thorne & Wheeler, Gravitation, Ch. 34 (global techniques, horizons and singularity theorems) · Wald, General Relativity, §9.2–9.5 (energy conditions, Raychaudhuri equation, singularity theorems) · Hawking & Ellis, The Large Scale Structure of Space-Time, §4.3, §8.2 · Carroll, Spacetime and Geometry, §4.6, §9.4 · Poisson, A Relativist's Toolkit, §2.1, §2.4 · Curiel, A Primer on Energy Conditions (2017) and Kontou & Sanders, Energy conditions in general relativity and quantum field theory (2020) for the modern averaged/quantum picture. Key results collected here for reference: the perfect-fluid reductions are NEC $\rho+p\ge0$ , WEC $\rho\ge0,\,\rho+p\ge0$ , DEC $-\rho\le p\le\rho$ , SEC $\rho+p\ge0,\,\rho+3p\ge0$ ; the geodesic-focusing bounds are $\Delta\lambda\le3/|\theta_0|$ (timelike) and $2/|\theta_0|$ (null); the logical chain is DEC $\Rightarrow$ WEC $\Rightarrow$ NEC and SEC $\Rightarrow$ NEC, with SEC and WEC independent; the Casimir energy density and pressure at $a=1\ \mu\mathrm{m}$ are $-4.3\times10^{-4}\ \mathrm{J/m^3}$ and $-1.3\ \mathrm{mPa}$ , from coefficients $-\pi^2/720$ and $-\pi^2/240$ ; the free-fall time at solar mean density is $t_{\rm ff}=\sqrt{3\pi/32G\rho}\approx1.77\times10^{3}\ \mathrm{s}$ ; and $\Lambda>0$ violates only SEC.

Builds on: Causal Structure of Spacetime · Null Geodesics and Light Cones · Conformal Transformations and Causal Invariance Leads to: Global Hyperbolicity · Domain of Dependence and Cauchy Horizons · Asymptotic Flatness

Ask the tutor: Ask me to trace any one energy condition from its inequality on $T_{\mu\nu}$ all the way to the theorem it proves — say, NEC to the area law, or DEC to positive mass — and we will work the Raychaudhuri focusing step by step.

What is the difference between an energy condition on matter and one on curvature?

Strictly, energy conditions are inequalities on the stress-energy tensor $T_{\mu\nu}$ — statements about matter. They become statements about curvature only after you invoke the Einstein equation, which converts NEC ( $T_{\mu\nu}k^\mu k^\nu\ge0$ ) into the null convergence condition ( $R_{\mu\nu}k^\mu k^\nu\ge0$ ). The distinction matters twice. The singularity theorems need only the curvature version, so they hold in any theory whose geometry focuses geodesics. And in modified gravity (say $f(R)$ ) the map between the two sides differs, so matter that looks exotic in Einstein gravity can be ordinary once the extra gravitational degrees of freedom are counted.

Why does the strong energy condition fail for the universe we actually live in?

Because the universe is accelerating, and acceleration requires $\rho+3p<0$ , the exact negation of SEC. The 1998 Type Ia supernova results showed the expansion is speeding up, which in the Friedmann equations means the dominant component has $p\approx-\rho$ — a cosmological constant or close to it. Since $\rho+3p=-2\rho<0$ there, SEC is violated on cosmological scales today. This is no crisis: SEC was always the most fragile of the four, and its failure is what "repulsive gravity" means. The consequence is that Hawking's cosmological singularity theorem, which assumes SEC, does not by itself force a Big Bang singularity in a dark-energy universe — one reason bouncing and inflationary alternatives remain live.

Do energy conditions forbid wormholes and warp drives outright?

Not outright — they make them expensive and fragile. A traversable wormhole throat must defocus light to stay open, the time-reverse of focusing, which requires $T_{\mu\nu}k^\mu k^\nu<0$ : a violation of NEC by "exotic matter". Because quantum fields can violate NEC pointwise (the Casimir vacuum is proof of principle), such geometries are not immediately ruled out. What constrains them are the averaged and quantum inequalities, which bound the total and duration of negative energy so tightly that a macroscopic wormhole needs implausibly fine-tuned states and the Alcubierre drive needs negative energies comparable to a galaxy's mass. The 2017 Gao–Jafferis–Wall construction made a genuinely traversable wormhole from a quantum coupling — but only a microscopic, briefly-open one, exactly the regime the inequalities permit.

How does Hawking radiation evade the area theorem?

The area theorem says a future event horizon's area never decreases, proved by applying Raychaudhuri to the horizon's null generators with NEC keeping their expansion non-negative. Hawking radiation shrinks the black hole, so it must break a hypothesis — and the one it breaks is NEC. The renormalized stress-energy near the horizon carries a negative-energy flux inward (the partner of the outgoing radiation), so $T_{\mu\nu}k^\mu k^\nu<0$ there, the generators focus, and the area can fall. This confirms rather than contradicts the logic: the classical theorem holds exactly as long as its energy condition does, and the one process that shrinks a black hole is the one that violates it. The generalized second law, $\mathrm{d}(S_{\rm BH}+S_{\rm matter})\ge0$ , restores monotonicity once the radiation's entropy is counted.

Why is the averaged null energy condition preferred over the pointwise one?

Because it is both survivable and sufficient. Pointwise NEC is violated by essentially every quantum field — Casimir vacua, squeezed states, the near-horizon region of an evaporating black hole all have locally negative energy — so building the theorems on it invites quantum corrections to knock them down. ANEC instead integrates $T_{\mu\nu}k^\mu k^\nu$ along a complete null geodesic and asks only that the total be non-negative, allowing local dips as long as they are compensated along the ray. Remarkably, ANEC can be proven from the axioms of quantum field theory, and it still powers topological censorship and averaged singularity and positive-mass theorems. It is the form in which "energy is mostly positive" survives quantization.

Is the dominant energy condition really about energy moving slower than light?

Yes, literally. DEC has two parts: the weak condition ( $\rho\ge0$ for every observer) plus the requirement that the measured flux $J^\mu=-T^\mu{}_\nu t^\nu$ be future-directed causal — never spacelike. A spacelike flux would mean energy streaming faster than light, so $J^\mu J_\mu\le0$ is a no-superluminal-transport rule. For a perfect fluid it tightens to $\rho\ge|p|$ : the pressure never exceeds the energy density. This is why DEC is the natural hypothesis for the positive-mass theorem — if every local energy flux is causal, the total mass at infinity cannot come out negative, and flat spacetime is stable.