Publication Type:
- Articles and Reports
Weapon Program:
- Nuclear
The U.S. and Israeli military strikes on Iran’s nuclear infrastructure in June set back Iran’s nuclear program, but by how much and for how long is uncertain. It is impossible to predict, at present, how long it would now take Iran to enrich enough uranium to fuel a small nuclear arsenal. However, one clear result is that such work is more likely than ever to take place at secret sites. Moreover, at least in the near term, Iran would likely seek to use its surviving stockpile of enriched uranium to kick-start a dash for a bomb.
The analysis below considers scenarios involving three sizes of secret sites and differing levels of usable enriched uranium stockpiles to establish theoretical estimates for how long it might take Iran to manufacture fuel for a small nuclear weapon arsenal in the aftermath of the strikes. Once the sites are set up and the centrifuge cascades installed, Iran could theoretically achieve this outcome in less than a month, assuming most material and some centrifuges are recoverable. However, it could take several years if little or none of the material and equipment Iran had accumulated before the strikes can be recovered.
In addition to enriching uranium to weapon-grade, some “weaponization” work would have to be done to craft a nuclear warhead. Iran also faces strategic and operational constraints, exacerbated by the strikes, that may influence Iranian leaders’ decision to restart the nuclear program.
Nuclear Weapon Potential of Iran’s Enriched Uranium Stockpile
Before the June strikes, Iran had accumulated a stockpile of enriched uranium sufficient to fuel more than two dozen nuclear weapons if enriched to weapons grade. That stockpile, comprised of 408.6 kg of uranium enriched to 60% fissile purity, 274.5 kg of 20% enriched uranium, and 5,508.8 kg of 5% enriched uranium, may now be temporarily inaccessible, but, according to Israeli Prime Minister Benjamin Netanyahu, it was not destroyed.
Iran would likely make use of this stockpile in a near-term dash for a nuclear arsenal because the alternative—enriching natural uranium—would be a time-consuming process that could carry a high risk of eventual detection.[1] Using uranium that has not already been converted to uranium hexafluoride gas (UF6) would also require the use of conversion facilities. Iran’s declared conversion facility at the Esfahan Nuclear Technology Center was heavily damaged in the June strikes, and the need to build a new conversion facility or activate an undeclared one would compound the risk of detection.
Reports vary as to the enriched uranium stockpile’s present location. Some or all of the material may have been moved before it could be bombed, or it may be buried by bombing debris. In the latter event especially, Iran may find it too risky or impractical to retrieve the full amount of the stockpile. The tables below therefore estimate the total number of weapons Iran could produce if it were to enrich to weapon grade either its full pre-strike stockpile or only half of it.
Table 1: Weapon Potential of Iran’s Pre-Strike Enriched Uranium Stockpile[2]
| Enrichment Level of Stockpiled UF6 | Quantity (kg) | Amount Needed for One Weapon (kg) | Number of Weapons |
|---|---|---|---|
| “Up to” 60% | 408.6 | 33.40 | 12 |
| “Up to” 20% | 274.5 | 101.36 | 2 |
| “Up to” 5% | 5,508.8 | 427.14 | 13[3] |
Table 2: Weapon Potential of Half of Iran’s Pre-Strike Enriched Uranium Stockpile
| Enrichment Level of Stockpiled UF6 | Quantity (kg) | Amount Needed for One Weapon (kg) | Number of Weapons |
|---|---|---|---|
| “Up to” 60% | 204.3 | 33.40 | 6 |
| “Up to” 20% | 137.2 | 101.36 | 1 |
| “Up to” 5% | 2,754.4 | 427.14 | 6 |
Timetable for Weapon-Grade Enrichment at Secret Sites
The U.S. intelligence community has long argued that secret sites pose the primary nuclear weapon threat from Iran because enriching uranium to weapon grade at hidden sites would minimize the likelihood of detection and attack. Now that an attack has actually occurred, and has destroyed or damaged Iran’s known sites, this argument has gained force. In any future dash for a nuclear arsenal in Iran, operational logic points toward doing so at undeclared sites, and particularly at sites with a smaller footprint that would be easier to hide.
Enrichment of Iran’s surviving uranium stockpile to weapon grade at a secret plant would require gas centrifuges to be installed at that site. As of late May, Iran was operating approximately 18,000 centrifuges across three uranium enrichment sites: two at Natanz and one, deeply buried, at Fordow. All three sites were damaged in the Israeli and U.S. strikes in June, and many of the centrifuges within them are presumed to have been destroyed either through direct impact or indirectly through the shock from vibration and/or sudden loss of power. However, centrifuges that were installed but not operating at the time of the strikes would have been less vulnerable to indirect impacts, and some may be salvageable. Iran had nearly 4,000 of its advanced centrifuge models (IR-2m, IR-4, and IR-6) in such a status at the end of May.[4]
Additionally, the IAEA has not been permitted to monitor Iranian centrifuge production since early 2021. Iran has therefore had a four-year window of opportunity to stockpile centrifuges for potential use in secret sites. Had Iran produced 10% more centrifuges than the number it installed at its declared sites between February 2021 and May 2025, it would have more than 1,300 unused advanced centrifuges on hand currently.[5]
The enrichment plant at Fordow, which was publicly exposed in 2009, was originally built clandestinely to house about 3,000 centrifuges. These centrifuges require only about 32,000 square feet, equal to approximately twice the size of the ice surface of a professional hockey rink.[6] A plant half that size would contain roughly 1,500 centrifuges and be approximately equivalent to the ice surface of one hockey rink. A secret enrichment facility of either size could be plausible given the abovementioned estimates of the number of centrifuges available to Iran.
The two tables below estimate the minimum theoretical time it would take Iran to produce sufficient fuel for five nuclear weapons of the implosion type[7] at secret enrichment sites containing 3,132 and 1,566 centrifuges, respectively. The estimates assume that the centrifuges are set up in 174-machine cascades of IR-2m, IR-4, and IR-6 centrifuges at a ratio of 4-3-2, which mirrors the cascade size and ratio of Iran’s installed centrifuges prior to the strikes.[8] Each table displays estimates of the time necessary to carry out the enrichment using 60% highly-enriched uranium (HEU), 5% low-enriched uranium (LEU), or natural uranium.[9] The time it would take Iran to build a secret enrichment plant and to install and prepare the centrifuges to operate is not accounted for in the below estimates.
Table 3: Estimated Minimum Time to Produce Fuel for Five Nuclear Weapons at a 3,132-Centrifuge Site
| Enrichment Level of Feedstock | Quantity (kg) | Enrichment Work (SWU)[10] | Time (days)[11] |
|---|---|---|---|
| HEU “up to” 60% | 167.0 | 560.5 | 15 (~2 weeks) |
| LEU “up to” 5% | 2,135.7 | 5,987.0 | 153 (~5 months) |
| Natural uranium | 22,425.0 | 19,463.8 | 496 (~1 yr. 4 mos.) |
Table 4: Estimated Minimum Time to Produce Fuel for Five Nuclear Weapons at a 1,566-Centrifuge Site
| Enrichment Level of Feedstock | Quantity (kg) | Enrichment Work (SWU) | Time (days)[12] |
|---|---|---|---|
| HEU “up to” 60% | 167.0 | 560.5 | 29 (~1 month) |
| LEU “up to” 5% | 2,135.7 | 5,987.0 | 305 (~10 months) |
| Natural uranium | 22,425.0 | 19,463.8 | 992 (~2 yrs. 9 mos.) |
Iran could instead build even smaller sites. An enrichment plant consisting of approximately 1,000 centrifuges would be roughly equivalent in size to two-and-a-half NBA basketball courts. As such a site may be easier to hide, Iran might decide to split its centrifuges among two or more locations in order to increase the program’s resiliency. Were one plant to be discovered and bombed, the other could continue to operate.
The following two tables estimate the minimum enrichment time for fuel for five implosion-type weapons using sites containing 1,044 centrifuges (six cascades) each, configured into three IR-2m cascades, two IR-4 cascades, and one IR-6 cascade. The first table contains estimates for a single site, and the second table contains estimates for two sites of that size working simultaneously. The time required for three such sites working simultaneously is comparable to the estimated time for a 3,132-centrifuge plant as shown in Table 3 above.
Table 5: Estimated Minimum Time to Produce Fuel for Five Nuclear Weapons at a 1,044-Centrifuge Site
| Enrichment Level of Feedstock | Quantity (kg) | Enrichment Work (SWU) | Time (days)[13] |
|---|---|---|---|
| HEU “up to” 60% | 167.0 | 560.5 | 45 (~1.5 months) |
| LEU “up to” 5% | 2,135.7 | 5,987.0 | 473 (~1 yr. 3.5 mos.) |
| Natural uranium | 22,425.0 | 19,463.8 | 1,535 (~4 yrs. 2.5 mos.) |
Table 6: Estimated Minimum Time to Produce Fuel for Five Nuclear Weapons at Two 1,044-Centrifuge Sites
| Enrichment Level of Feedstock | Quantity (kg) | Enrichment Work (SWU) | Time (days)[14] |
|---|---|---|---|
| HEU “up to” 60% | 167.0 | 560.5 | 23 (~3 weeks) |
| LEU “up to” 5% | 2,135.7 | 5,987.0 | 237 (~8 months) |
| Natural uranium | 22,425.0 | 19,463.8 | 768 (~2 yrs. 1 mo.) |
Weaponization
For uranium enriched to weapon grade to be usable in a nuclear weapon, it would have to be processed further into weapon components. Also, the other parts of a successful weapon would have to be ready to receive the uranium. Fabricating these other components could be done in parallel with uranium enrichment and could take place on a laboratory scale, which would make it difficult to detect.
The June strikes have likely hindered Iran’s ability to complete the weaponization process quickly. Prior to the strikes, estimates of the time required ranged from a few months to two years. However, the Israeli and U.S. attacks damaged facilities in the Esfahan Nuclear Technology Center that Iran has used or planned to use to convert UF6 into uranium metal, an essential step in weaponization.[15] Israeli strikes also killed several senior Iranian nuclear scientists who had been involved in nuclear weapon-related research.
Nonetheless, Iran would not be starting from scratch with weaponization. According to an investigation by the IAEA into "possible military dimensions" of Iran's nuclear program, Iran had a coordinated nuclear weapon program between 1999 and 2003. Specifically, the IAEA found that Iran developed several components of a nuclear weapon and undertook related research and testing. Information obtained by Israeli intelligence and revealed in April 2018 indicated that Iran sought to preserve this program after 2003 by dividing its nuclear program between covert and overt activities and retaining an expert team to continue work on weaponization. In mid-2024, the U.S. intelligence community dropped its longstanding assertion that “Iran is not currently undertaking the key nuclear weapons-development activities necessary to produce a testable nuclear device.”
The analysis in the sections above also assumes that Iran would use 16 kg of highly enriched uranium metal (about 90% U-235) in the finished core of each nuclear weapon. Sixteen kilograms are assumed to be sufficient for an implosion weapon. This was the amount called for in a design for such a device that has circulated on the nuclear black market, to which Iran has had access.
Some experts believe that Iran could use less material (as few as seven kilograms of over 90% enriched uranium), assuming Iran would accept a lower yield for each weapon.[16] Conversely, Iran could use more material at a lower level of enrichment (60%) and rely on a “gun type” nuclear weapon such as the one dropped on Hiroshima. Iran may still have access to enough material for one or more such devices, which would not require testing, but which would be heavier and larger than an implosion device.
Other Variables
In addition to the weapon potential described above, there are other factors that could powerfully influence any decision by Iran to restart its nuclear effort. One is its military weakness. Without effective air defenses in the wake of the June strikes, Iran’s nuclear sites and other targets such as its oil industry are vulnerable to severe damage at any time. Another is its lack of operational security. The June strikes proved that Israeli intelligence had thoroughly penetrated Iran’s nuclear program and much else. That fact could make it risky for Iran to try to hide centrifuges or enriched uranium, or to restart an enrichment program at secret sites.
A third constraint is Iran’s loss of nuclear personnel. With its top levels of scientists and engineers killed, and Israel reportedly threatening to kill their replacements, Iran may find it difficult to staff a revived program.
Footnotes:
[1] The risk of detection is assumed to be a function of conspicuousness and time. A conspicuous activity risks attracting notice quickly; a time-consuming activity risks eventually being discovered, on the assumption that each day carries a certain risk of detection through some development such as a project insider turning informant or an operational security mishap that unintentionally reveals the activity. A covert operation of this type would seek to be both inconspicuous and swift.
[2] This table uses IAEA estimates of Iran’s stockpile as of May 17, 2025. These theoretical calculations are generated using a SWU calculator published by URENCO, a European uranium enrichment consortium. They assume that 20 kg of 90% U-235 in the form of UF6 would be needed for each weapon. The tails are assumed to be 0.3% and because the IAEA describes the enrichment level as "up to" a percentage, a lower feed enrichment percentage than the stated upper limit (54% for uranium enriched “up to” 60%, 18% for uranium enriched “up to” 20%, 4.5% for uranium enriched “up to” 5%) is used for these calculations. For comparison, natural uranium is about 0.7% U-235.
[3] Includes the use of the remaining 7.8 kg of 60% uranium and 71.78 kg of 20% uranium not yet accounted for in the first two figures.
[4] On the eve of the war, Iran was operating an estimated 5,666 IR-2m centrifuges out of 7,060 installed, leaving 1,394 for possible salvage. Iran was operating 2,489 out of 4,321 IR-4 centrifuges, leaving 1,832 for possible salvage. Finally, Iran was operating 2,552 out of 3,256 IR-6 centrifuges, leaving 704 for possible salvage. Those potentially salvageable advanced centrifuges together make up a total of 3,930 machines. Were approximately three-quarters of those machines to be salvaged, Iran could use them to establish an enrichment plant roughly the size of Fordow.
[5] In February 2021, when monitoring stopped, Iran had installed or was in the process of installing 690 IR-2m centrifuges, 300 IR-4 centrifuges, and 302 IR-6 centrifuges at its declared sites (See: GOV/2021/10 and GOV/INF/2020/10). By May 2025, it had an estimated 7,060 IR-2m centrifuges, 4,321 IR-4 centrifuges, and 3,256 IR-6 centrifuges installed at declared site. This means that Iran likely produced at least 6,370 IR-2m centrifuges, 4,021 IR-4 centrifuges, and 2,954 IR-6 centrifuges during the past four years. If Iran produced and stored ten percent more centrifuges of each type, that would amount to 637 IR-2m centrifuges, 402 IR-4 centrifuges, and 295 IR-6 centrifuges, for a total of 1,334 additional machines.
[6] Each centrifuge is assumed to require about one square meter (10.7 square feet) of space, the amount used in Iran’s enrichment plant at Natanz. The ice surface of a National Hockey League rink is 200 feet long and 85 feet wide.
[7] Under the Amad Plan, Iran’s pre-2004 clandestine nuclear weapons effort, Iran had a goal to design, produce and test five warheads, each with 10 kiloton TNT yield, for integration on a ballistic missile.
[8] IR-1 centrifuges are assumed not to be used in this scenario, as Iran’s large-scale production of more advanced centrifuges has rendered the IR-1 largely obsolete. IR-1 centrifuges would also not be well-suited for secret sites, as their lower efficiency means that more centrifuges (and therefore a larger site) would be required to achieve a given level of enrichment. Iran has yet to produce its other advanced centrifuge models—IR-5, IR-6s, IR-7, IR-8, IR-9—at scale.
[9] Uranium enriched up to 60% and 5% comprised the largest portions of Iran’s pre-strike HEU and LEU stockpiles, respectively. Halving those stocks, as done in the Weapon Potential section above, would not make a difference to this calculation. Half of Iran’s 408.6 kg 60% uranium stockpile, 204.3 kg, would still be sufficient for five weapons. The same is true for Iran’s 5% uranium stockpile.
[10] The separative work unit (SWU) is the standard measure of the effort (work) required to increase the concentration of the fissionable U-235 isotope. According to pre-2016 production data from Natanz, Iran's IR-1 centrifuges have achieved an average annual output of about .8 separative work units, or SWUs, per machine. The IR-2m and IR-4 centrifuges are based on Pakistan's P-2 centrifuge and is assumed in these estimates to have an operational output of 4 SWU (See: Alexander Glaser, "Characteristics of the Gas Centrifuge for Uranium Enrichment and Their Relevance for Nuclear Weapon Proliferation (corrected)," Science and Global Security, Vol. 16, Nos. 1-2 (2008), p. 9). The IR-6 is assumed in these estimates to have an operational output of 6.6 SWU. At Fordow in 2022, Iran used two cascades of IR-6 machines enriching natural uranium to produce the 5% feed for the IR-1 centrifuges enriching up to 20% U-235 (See: GOV/2022/62 Paras. 27-29). To produce enough feed for this configuration, each IR-6 machine would have to produce at least 6.6 SWU.
[11] Eight cascades of IR-2m centrifuges (1,392 machines) would generate about 5,568 SWU per year assuming an operational capacity of 4 SWU per machine. Six cascades of IR-4 centrifuges (1,044 machines) would generate 4,176 SWU per year assuming an operational capacity of 4 SWU per machine. Four cascades of IR-6 centrifuges (696 machines) would generate 4,593.6 SWU per year assuming an operational capacity of 6.6 SWU per machine. In total, the cascades operating in production mode at a secret site equipped with 3,132 centrifuges in the abovementioned configuration could generate up to about 14,337.6 SWU annually, or about 39.28 SWU per day. See footnote 10 for assumptions related to the operational SWU output of Iran’s centrifuge models and footnote 2 for assumptions related to product, tails, and feedstock.
[12] Four cascades of IR-2m centrifuges (696 machines) would generate about 2,784 SWU per year assuming an operational capacity of 4 SWU per machine. Three cascades of IR-4 centrifuges (522 machines) would generate 2,088 SWU per year assuming an operational capacity of 4 SWU per machine. Two cascades of IR-6 centrifuges (348 machines) would generate 2,296.8 SWU per year assuming an operational capacity of 6.6 SWU per machine. In total, the cascades operating in production mode at a secret site equipped with 1,566 centrifuges in the abovementioned configuration could generate up to about 7,168.8 SWU annually, or about 19.64 SWU per day. See footnote 10 for assumptions related to the operational SWU output of Iran’s centrifuge models and footnote 2 for assumptions related to product, tails, and feedstock.
[13] Three cascades of IR-2m centrifuges (522 machines) would generate about 2,088 SWU per year assuming an operational capacity of 4 SWU per machine. Two cascades of IR-4 centrifuges (348 machines) would generate 1,392 SWU per year assuming an operational capacity of 4 SWU per machine. One cascade of IR-6 centrifuges (174 machines) would generate 1,148.4 SWU per year assuming an operational capacity of 6.6 SWU per machine. In total, the cascades operating in production mode at a secret site equipped with 1,044 centrifuges in the abovementioned configuration could generate up to about 4,628.4 SWU annually, or about 12.68 SWU per day. See footnote 10 for assumptions related to the operational SWU output of Iran’s centrifuge models and footnote 2 for assumptions related to product, tails, and feedstock.
[14] Each plant would generate about 12.68 SWU per day (see footnote 12); the two plants working simultaneously would therefore generate about 25.36 SWU per day. See footnote 10 for assumptions related to the operational SWU output of Iran’s centrifuge models and footnote 2 for assumptions related to product, tails, and feedstock.
[15] In August 2021, Iran informed the IAEA that it had produced 200 grams of uranium metal enriched to 20% purity at the Fuel Plate Fabrication Plant (FPFP) at the Esfahan Nuclear Technology Center (ENTC) (See: GOV/INF/2021/39). Iran had subsequently begun construction of facilities in ENTC for the production of natural and enriched uranium metal. All of these facilities were damaged in the June strikes, according to the IAEA.
[16] See Thomas B. Cochran and Christopher E. Paine, “The Amount of Plutonium and Highly Enriched Uranium Needed for Pure Fission Nuclear Weapons,” (Washington, DC: Natural Resources Defense Council, revised April 13, 1995).