This guest post is by Carl Parkin. Jon Gambrell has a story with the Associated Press on the reconstitution of Iran’s missile program in the wake of the 12-Day War. That story is based in part on this report.
Introduction
At the outset of Operation Rising Lion, Israeli Prime Minister Benjamin Netanyahu justified Israel’s attack on Iran as not only a response to its nuclear weapons program, but also its significant ballistic missile capabilities. He called the threat that Iran’s missile program posed to Israel “existential” and stated that, without interference, Iran would acquire 20,000 missiles that could reach Israel by 2031. Later, Yonah Jeremey Bob at The Jerusalem Post reported details of Israeli intelligence estimates of Iran’s missile production, indicating an approximate production rate of 240 medium range ballistic missiles (MRBMs) per month, 2400 MRBMs per year. Bob also stated that the estimates hinged on the completion of a mysterious new Iranian facility. In early June however, Barak Ravid at Axios published a significantly more conservative production estimate of 50 ballistic missiles per month, citing a U.S. intelligence official.
The difference between these two estimates is dramatic and deserves investigation. We’ve done so by developing a methodology to estimate Iran’s solid propellant missile production capacity. Since Iran’s liquid-fuelled production line likely only marginally contributes to their total production capacity, we believe our solid propellant missile production estimates paint a compelling picture of their entire missile program. Through our estimate we believe we’ve identified why Israel and the US have such dramatically different perceptions of Iran’s missile production, and pinpointed the mysterious facility on which Israel’s latest projections were based.
Our work indicates that Israel’s estimate of Iran’s missile production capacity was reasonable, and that Iran could have been producing as many as 217 missiles per month if they were operating at their maximum capacity. However, while our analysis supports Israeli claims regarding missile production, it casts others regarding the “destruction” of the Iranian missile industry into doubt. The newest and largest Iranian solid propellant production plant at Shahroud — that mysterious facility mentioned earlier — was left largely untouched during Operation Rising Lion. Furthermore, in its strikes on Iran’s solid propellant production facilities, Israel has mainly targeted one element of the production process: mixing. The IDF appears to believe planetary mixers, a key component of solid propellant production, were the bottleneck in Iran’s medium-range ballistic missile program. If Iran is capable of re-acquiring mixers quickly, the Islamic Republic could feasibly return to a production capacity matching Israel’s assessments given the lack of damage to other parts of the solid propellant production process. However, if mixing wasn’t the bottleneck in production, it’s possible that they could be back at full capacity even sooner.
It is worth noting that Israel also struck other elements of missile production, including guidance kit assembly buildings and carbon fiber production plants. While motor production may be quickly rebuilt, it’s possible these other elements could continue to bottleneck Iran’s recovery.
Summary of Methodology & Findings
To explain how we reached our estimate and conclusions, we must explain how solid propellant missile production works. Iran’s solid rocket motors (SRMs) are made from a composite propellant. To make composite propellants fuel and oxidizer, both in powdered forms, are mixed with an epoxy that binds them together into a homogenous slurry. This slurry is then poured, or “casted” in a rocket motor casing and turned, through continuous heating or “curing”, into a solid. Fit that finished motor with a nozzle, control section, warhead, and guidance, and you have a missile. For those who wish to know more about the process, we’ve found this video highly educational.
For the large SRMs that one needs for ballistic missiles or space launch vehicles, casting and curing are completed in large, concrete-walled, temperature controlled, underground pits. Our basic methodology for estimating Iranian SRM production is to determine the number of casting pits Iran has and how long the motors take to cure. Multiplying these two variables together should get us an estimate of how many motors Iran can make in a month.
Casting Pits X (31 / Curing Time) = Motors Per Month
Curing is the longest step of the missile production process, and other steps can likely catch up during the extended time it takes.1 Therefore, the amount of motors that can be cured in a given period is likely a direct analog for Iran’s solid propellant missile production rate.
Determining the total number of casting pits in Iran is relatively simple. The large concrete chutes required for casting are constructed before the actual buildings themselves, so casting facilities can be identified in satellite images taken during a building’s construction.
Iran has three known solid rocket motor production facilities: Khojir, Parchin, and Shahroud. We’ve counted the concrete chutes at each of these facilities, and from that count, derived a range of estimates for how many casting pits Iran has in total: between 44 and 56. Satellite imagery showing the concrete pits in question and more information on how we reached our estimate can be found below.
The remaining variable in our equation is the time that it takes Iran’s SRMs to cure. After consulting a number of papers about the curing properties of the solid propellant binder that Iran uses, Hydroxyl-Terminated Polybutadiene (HTPB), we’ve assessed that Iran’s SRMs take between 6 and 10 days in the casting pits to finish curing.
Plugging the different pit number and curing time estimates into our equation, we get a range of estimates for Iran’s missile production capacity: between 136 and 289 solid propellant missiles per month. This is a pretty wide range, but we can offer a more refined one that we believe is more accurate. Excluding some of the pits we judge to be too small to produce missiles that can reach Israel, and cutting out the shortest cure time of 6 days which was less represented in the literature on curing times, we get a more narrow range of 136 to 217. We believe that the highest estimate within this range, 217 missiles per month, is an effective recreation of the Israeli estimate of 240 missiles per month. The minor difference between our high-end estimate and Israel’s can be explained by adding liquid propellant missile production, as the 217 missiles were only solids.
To explain the United States’s estimate of 50 missiles per month, we can make some reasonable modifications to our variables. Iran’s newest SRM production line at Shahroud only began construction in 2024; this is likely the unidentified facility expansion that the Jerusalem Post claimed Israel’s estimate relied on. Given that the United States was estimating current production rather than future production as Israel did, we can assume the U.S. estimate did not incorporate the casting pits at these most recent facilities. This leaves us with a dramatically lower pit range of 28 to 31. Additionally, our curing time estimate is the weakest of our assumptions, given that it is largely based on studies involving small amounts of propellant, in some cases only a handful of millimeters thick. While these studies explain that curing time is only somewhat dependent on sample size, the U.S. may believe the larger diameters of solid rocket motors have a more significant impact on, and therefore use a longer curing time estimate of 10-14 days rather than our range of 6-10 days.
These modified ranges leave us with a new range of estimates for missiles per month: between 62 and 95. This range is still higher than the U.S. estimate, even without incorporating Iran’s liquid propellant program. This discrepancy indicates that the U.S.’s estimate is of real production rate, rather than the maximum production capacity provided by casting facilities. Perhaps the U.S. intelligence community believes Iran has made an intentional decision not to produce at maximum capacity, or believes that Iran is limited by other steps in the production process, such as propellant precursor procurement.
The following table outlines the major components of our estimate, and how it differs when attempting to replicate the Israeli or U.S. numbers. The leftmost columns identify different concrete chutes in which casting pits are constructed across Iran and their respective sizes. We then represent our range of estimates for how many pits can be constructed in each chute, getting us to our total number of each pit type. Finally, we use our range of days to cure (6-10 for Israel, 10-14 for the US) to calculate a final range of motor production estimates for Israel and the US, omitting the latest pits at Shahroud for the US estimate.
| Type | Pit Types | Start Date | Chutes | Length (m) | Width (m) | Pits per chute | Days to cure (low Israel) | Days to cure (high Israel, low US) | Days to cure (high US) | Israeli Projection | US Range (2025) |
| SH-A1 | Shahroud Original (Big) | 2011 | – | – | – | 1 | 6 | 10 | 14 | 3-5 | 2-3 |
| SH-A2 | Shahroud Original (Middle) | 2011 | – | – | – | 1 | 6 | 10 | 14 | 3-5 | 2-3 |
| SH-A3 | Shahroud Original (Small) | 2011 | – | – | – | 1 | 6 | 10 | 14 | 3-5 | 2-3 |
| SH-B | Shahroud Standard | 2017/2024 | 25 | 5.2 | 4.3 | 1 | 6 | 10 | 14 | 78-129 | 27-37 |
| SH-C | Shahroud Wide | 2024 | 3 | 5.2 | 6.18 | 1-2 | 6 | 10 | 14 | 9-31 | – |
| SH-D | Shahroud Small | 2024 | 6 | 2.1 | 1.9 | 0 | 6 | 10 | 14 | 0-31 | – |
| KH-A | Khojir One | 2014 | 1 | 3.25 | 4 | 1 | 6 | 10 | 14 | 3-5 | 2-3 |
| KH-B | Khojir Two | 2020 | 1 | 5.5 | 12.5 | 3 | 6 | 10 | 14 | 9-6 | 7-9 |
| KH-C | Khojir Three | 2020 | 1 | 3.5 | 10 | 3-4 | 6 | 10 | 14 | 9-21 | 7-12 |
| PA-A | Parchin | 2021 | 2 | 12.5 | 6.5 | 3-4 | 6 | 10 | 14 | 19-41 | 13-25 |
| Total Missiles | 136-289 | 62-95 |
The remainder of this article is dedicated to a more systematic exploration of our methodology. First we’ll dive into how we counted casting pits at Iran’s various SRM production facilities.
Counting Casting Pits: Shahroud
Let’s start with Shahroud. Shahroud is the largest facility in terms of casting capacity. Its sprawling footprint contains multiple separate solid rocket motor production lines. The first was built when the plant was originally constituted, circa. 2011. Ground was broken on the second in 2017, along with a small production line to the Northeast, and construction began on the third and final in the spring of 2024.
Shahroud was attacked twice by Israel: once in October 2024 during Operation Days of Repentance, and again during Operation Rising Lion. During the October attacks, one of the buildings in the first production line was struck. Based on construction photos, the area surrounding this building has three casting pits.
Later additions to the plant increased its production capabilities. Three new casting buildings were added in the second production line, each housing four concrete chutes, for a total of twelve. Only one of these facilities was damaged in the October strikes — likely collateral damage from a neighboring mixing building which was directly targeted and destroyed.
Finally, the newest production line contributed a whopping 18 concrete chutes to the plant’s total capacity: three buildings, each with six chutes. Six of those 18 chutes, however, are about a quarter of the size of the others. These may be used for smaller missiles with shorter ranges, which are unlikely to have been included in Israel’s estimate. Our low end estimate will exclude these six pits, and our high end will include them. The October attacks left this portion of the facility largely untouched.
Finally, in the same phase of construction, an additional casting building was added in the smaller production line to the North-east of the plant. While its individual concrete chutes are more difficult to identify, we’re largely confident this one also has four because the concrete bank is 20 meters long and 5 meters wide — enough to fit four of the roughly 5 meter-wide chutes we see elsewhere in Shahroud.
Adding up the casting pits and the concrete chutes, Shahroud has 37 spaces for casting, 22 of which were constructed within the past year. The dramatic recent increase in this plant’s production capacity indicates that this facility was likely the one mentioned by Israeli intelligence, on which their large estimates were hinged.
Counting Casting Pits: Khojir and Parchin
Next up are Khojir and Parchin. Khojir has two buildings that are likely used for casting missiles that can reach Israel.2 One contains one, straightforward chute, roughly square in dimension, like the concrete areas at Shahroud. The other building is more complicated — it contains two concrete chutes that are especially large, and vary widely in dimension.
We encounter a similar situation at Parchin. There are two facilities with concrete chutes, but these chutes are large, taking up a solid portion of the buildings that they occupy.
Our assessment is these areas each contain multiple casting pits. This conclusion is based on these pits’ unusual characteristics relative to the more ‘standard’ pits at Shahroud, and an informed assumption on the layout of Iranian casting facilities.
We believe Iran’s casting facilities follow a layout seen at many other SRM production sites. Around the world, casting pits are consistently housed inside of recessed, square, concrete areas. This layout can be seen at two different pits at Northrop-Grumman’s Promontory plant in Utah.
And at a Regulus plant in Kourou, French Guiana.
We judge that this layout is also used in Iran for a number of reasons. Firstly this circular layout is well-suited for solid rocket motor casting and curing as it ensures even heat transfer during the curing process. This is necessary to make sure the propellant within each motor doesn’t crack or otherwise acquire flaws. Additionally, circular openings are easier to vacuum seal, a requirement of some casting processes. Finally, the rectangular area outside of the pit can serve as a dedicated work area and provides an underground space for the extensive wiring and piping that casting pit monitoring and heating requires. This leads us to believe that, rather than using the entire rectangular area seen in construction imagery for casting, the Iranians construct a casting pit — or multiple casting pits — within each area, and use the concrete space around it for wiring, piping, and dedicated work.
The concrete chutes at Khojir and Parchin are significantly larger than the others we’ve seen at Shahroud, and are rectangular, as opposed to the essentially square chutes we’ve seen elsewhere. We therefore assess that the irregular-sized concrete areas at Khojir and Parchin likely include multiple pits, with the rectangular shape providing space for a line of pits of a similar diameter. It’s difficult to provide well-informed estimates of how many pits each area could contain, but we can give a range. The table below shows how pit diameters would vary depending on how many pits the Iranians placed in each chute. The higher the number of pits, the lower the maximum pit diameter, and vice versa. It assumes a gap of .5 meters between the wall of the concrete area and the wall of the actual casting pit is required for wiring and workspace.
| Concrete Area | Number of Pits | Expected Pit Diameter |
| Parchin | 2 | 5.5 |
| 3 | 3.5 | |
| 4 | 2.5 | |
| 5 | 1.9 | |
| Khojir Smaller | 2 | 4.25 |
| 3 | 2.66 | |
| 4 | 1.875 | |
| 5 | 1.4 | |
| Khojir Larger | 2 | 5.5 |
| 3 | 3.5 | |
| 4 | 2.5 | |
| 5 | 1.9 |
Some of these pit diameters are less likely to have been chosen for construction. It’s unlikely that the Iranians would build pits that are larger than the most common pit at Shahroud, given that it’s the largest pit we’ve previously observed that is a highly likely candidate for medium-range ballistic missile production. It’s also unlikely that Iran would build pits that are smaller than their smallest likely medium-range ballistic missile pit — the smallest original pit at Shahroud.3 By excluding pit diameters that are bigger or smaller than these pits (those outside the range of 3.3-2.5m), we can narrow our range of expected pits per chute. Here’s the previous chart with the new bounds applied, along with some room for error to incorporate nearby pit diameters:
| Concrete Area | Number of Pits | Expected Pit Diameter |
| Parchin | 2 | 5.5 |
| 3 | 3.5 | |
| 4 | 2.5 | |
| 5 | 1.9 | |
| Khojir Smaller | 2 | 4.25 |
| 3 | 2.66 | |
| 4 | 1.875 | |
| 5 | 1.4 | |
| Khojir Larger | 2 | 5.5 |
| 3 | 3.5 | |
| 4 | 2.5 | |
| 5 | 1.9 |
This bounding allows us to reach our final pit estimates. The image below displays each different chute we can see in satellite imagery, and the subsequent table lays out how many pits we think are in each chute, and then uses those numbers to create high and low estimates for total pits.
| Legend | Pit Types | Chutes | Pits per chute | Pits (Low) | Pits (High) |
| Shahroud | |||||
| 1 | Shahroud Original (Big) | – | 1 | 1 | 1 |
| 2 | Shahroud Original (Middle) | – | 1 | 1 | 1 |
| 3 | Shahroud Original (Small) | – | 1 | 1 | 1 |
| 4 | Shahroud Standard | 25 | 1 | 25 | 25 |
| 5 | Shahroud Wide | 3 | 1-2 | 3 | 6 |
| 6 | Shahroud Small | 6 | 0-1 | 0 | 6 |
| Totals: | 31 | 40 | |||
| Totals without new production line (US): | 15 | 15 | |||
| Khojir | |||||
| 7 | Khojir One | 1 | 1 | 1 | 1 |
| 8 | Khojir Two | 1 | 3 | 3 | 3 |
| 9 | Khojir Three | 1 | 3-4 | 3 | 4 |
| Totals: | 7 | 8 | |||
| Parchin | |||||
| 10 | Parchin | 2 | 3-4 | 6 | 8 |
| Totals: | 6 | 8 | |||
| Totals Overall: | 44 | 56 | |||
| Totals US: | 28 | 31 | |||
Starting with Shahroud, we have the original 3 pits (1, 2, and 3). Next are the 25 standard-sized chutes across the plant’s different production lines (4), each with one pit. Then we can add the 1-2 pits that may be housed in the newest production line’s 3 slightly wider chutes (5). Finally, the 6 smallest pits at Shahroud (6) may or may not produce missiles that can hit Israel, so while they likely contain pits, we’ve given them an effective pit-per-chute range of 0-1. Adding up all these numbers gives us a range of 31 to 40 pits at Shahroud.
As for Khojir, its smallest, standard-sized chute (7) likely contains 1 pit. Its next largest chute (8) likely contains 3 pits, and the largest chute at the complex (9) houses 3-4. Adding these up gives us a total of 7-8 pits at Khojir. At Parchin, each of its two chutes (10) likely contains 3-4 pits, giving us a total of 6-8 pits at the site. With Khojir, Parchin, and Shahroud added up, our total range of pits is between 44 and 56. This is the range that we used to replicate Israel’s estimates. We then narrowed our range to 44-50 for our second, more accurate reproduction of Israeli numbers, as we doubt that the 6 small pits at Shahroud would be incorporated in their estimates. Finally, to reach the U.S’s estimates, we subtracted all of the pits from the newest facility expansion at Shahroud — twelve standard chutes (4), and all of the wide (5) and small (6) chutes, leaving us with a pit range of 28 to 31.
Curing Times
The final variable in our equation is the time it takes each motor to cure. Recall that solid propellant slurry is made up of three ingredients: fuel, oxidizer, and binder. Curing is a chemical reaction: the binder’s polymer chains ‘cross-link’ in response to heat, bonding together and changing the mixture from a slurry into a dense, hard solid. Curing time varies depending on the binder that is used; the Iranians use Hydroxyl-Terminated Polybutadiene, or HTPB. A paper studying the effect of curing times and temperatures on the properties of cured HTPB-bound propellants names a standard curing time of 6 days at 60 degrees celsius. Two other papers involving the preparation of HTPB samples name a curing time of 7 days at a variety of temperatures. U.S. and Chinese studies of HTPB propellant aging prepared their samples by curing for 10 days. Finally, an Iranian study produced by faculty at the Defense Ministry’s Malek Ashtar University of Technology, models HTPB curing times based on a variety of factors, including aspect ratio of the sample (length relative to width). When discussing cylinders with length exceeding four times their width, it approximates curing time as 7.2 days. Exact curing time varies with diameter, but the sheer volume of studies indicating a curing time between six and ten days indicates that such variance is not incredibly significant. This leaves us with the 6 to 10 day range that we used for the Israeli estimate. As previously stated however, a majority of these papers are discussing the curing of relatively small amounts of HTPB, and do not take into account the extra time it would likely take to cure a sample with a larger diameter. Therefore, we think it is fair to increase the expected curing time to 10-14 days when attempting to replicate the United States’s more conservative estimate.
Addendums & Conclusion
Our final motor production estimates come with a disclaimer, which is seemingly unaccounted for by both the U.S. and Israel. There are sites in the U.S. and elsewhere where multiple motors are casted and cured in the same pit. The image below is from Promontory showing such an arrangement.
It’s difficult to say whether the Iranians do this. Multiple motors per pit may require a more powerful heating apparatus, considering the additional surface area and material involved. There may be other technical barriers to this sort of process which are difficult to determine. But, if we start thinking in terms of multiple motors per pit, our estimates begin to rise far above those of even the Israelis. Even two motors per pit would double our estimates.
However, there is some reason to think this might be possible: our projections of Iran’s casting pit diameters are quite large. Approximating from the pits at Promontory and Kourou, we can estimate that the ‘wiggle room’ required between the pit and the edge of the concrete area for wiring, piping, and workers moving around is about .5 meters. With that in mind, the standard pits at Shahroud could be as big as 3.3 meters. Iran’s solid propellant medium-range ballistic missiles are much smaller in diameter: the Dezful’s is .68 meters, and the Kheibar Shekan and Fattah-1 are both .76 meters across. Several of these motors could feasibly fit in the pits if they are actually that big. However, it’s entirely possible the ratio of concrete area to pit diameter differs significantly between the Western sites and sites in Iran — while technical requirements and common sense lead us to believe that the Iranian layout similarly consists of circular pits in a recessed, rectangular concrete area, there is no convincing technical reason that the space between that area and the pit’s edge should be .5 meters instead of any other distance. Additionally, the Iranians continue to build pits of different sizes, demonstrated by the two differently-sized types of pits at the newest production line at Shahroud. It’s difficult to explain why they would build pits of different specifications if they could cast smaller motors in big pits: why not build a big pit for flexibility’s sake?
With all of that in mind, we think our estimate paints a solid picture of why U.S. and Israeli estimates differ so wildly. While Netanyahu’s rather ambitious statements about the Iranian missile program warranted investigation, we’ve found there’s good reason Israeli intelligence would estimate such extreme figures.
However, we’ve also concluded that Israel’s strike on Iran’s missile program during Operation Rising Lion had significant limitations. The attacks only destroyed a few solid propellant production buildings, namely those that housed mixers. And Israel only struck one small support building in Shahroud’s latest production line — a strange choice given that this expansion was used to justify the attack in the first place.
Israel’s targeting indicates that they believed mixing was a bottleneck in Iran’s missile production. However, by focusing on this perceived bottleneck, they’ve left the casting portion of the facilities untouched. Therefore, if Iran is able to overcome their mixing limitations, they’ll have all the casting capacity that they need to start producing at high volumes again.
And, going off of the analysis done by CNS Research Associate Sam Lair and reporting from the Wall Street Journal, Israeli and U.S. missile defense interceptors were measurably depleted by Iran’s attacks in June. When Iran’s solid propellant missile plants are fully operational again, will Israel and the U.S. be able to mount as effective a defense as they have prior? Only time will tell.
Carl Parkin is a Summer Undergraduate Fellow at the Center for Nonproliferation Studies, working on the New Tools team. He is also a Foreign Policy Futures Fellow with the Reimagining U.S. Grand Strategy Program at the Stimson Center, and a rising senior at Franklin & Marshall College.
- Alain Davenas, “Composite Propellants,” in Solid Rocket Propulsion Technology, ed. Alain Davenas (Oxford: Pergamon, 1993), 415–75, https://doi.org/10.1016/B978-0-08-040999-3.50015-1.
- A second solid rocket motor production line exists at Khojir (35.683° N, 51.654° E), but is likely used for short-range ballistic missiles like the Fattah-110, since it was constructed long before Iran ever debuted its first solid propellant medium-range ballistic missile.
- The largest pit at Shahroud, at the original production line, has a diameter of 5.5 meters. We assess that this pit is likely an anomaly, constructed as part of Iran’s first experiments with large solid propellant production at Bidganeh and Shahroud which were tied to long-range ballistic missiles and space launch vehicle construction. We doubt that pits of this size would be constructed at a missile production line. Additionally, we’ve opted to use the smallest original pit at Shahroud rather than its newest small pits, since we believe the newest pits are likely used for shorter range missiles.
