SpaceX Super Heavy / Starship Environmental Assessment Report

We can lick gravity, but sometimes the paperwork is overwhelming.
— Wernher von Braun

SpaceX is in the process of developing a completely reusable two-stage super-heavy class orbital launch vehicle called Super Heavy / Starship.  This is the latest iteration in an evolving design which has previously been called the Mars Colonial Transporter, Interplanetary Transport System, and Big Falcon Rocket (BFR).  The present design (which continues to evolve) specifies a payload to low Earth orbit (LEO) between 100 and 150 tonnes.  This compares to the 140 tonnes to LEO of the Saturn V which was, of course, completely expendable.  The Super Heavy/Starship will be, if built to the current design, the largest and most powerful rocket ever, with a lift-off thrust of 62 meganewtons (MN), compared to 35.1 MN for the Saturn V.

On August 1st, 2019, NASA released the “Draft Environmental Assessment for the SpaceX Starship and Super Heavy Launch Vehicle at Kennedy Space Center (KSC)” [PDF, 21 Mb, slow to download].  This is a 250 page document, including supplementary material prepared by contractors, which assesses the environmental impact of the construction of facilities at the the Kennedy Space Center in Florida to launch Super Heavy/Starship and operational launches of the vehicle, building up to the planned launch cadence of 24 flights per year.

There is substantial new detail in this report which I’ve not seen before.  For example, the current plan is to land the Super Heavy booster on a drone ship at least 20 nautical miles off the Florida coast.  Earlier plans had it returning to the launch site and landing directly on the launch mount.  There are detailed maps of expected noise levels from launches, landings, and the sonic booms created during these operations.  They also consider a great many other things, including:

  • Cultural Resources
  • Air Quality
  • Climate
  • Socioeconomics
  • Environmental Justice and Children’s Environmental Health and Safety

The “Biological Resources” section is one of the longest, in which you will learn more about the sex life of the leatherback sea turtle (Dermochelys coriacea) than you probably wanted to know.

The reasonable and prudent measures and non-discretionary terms and conditions identified in the BO for CCAFS are similar to those for KSC to aid in the reduction of lighting impacts and incidental take of sea turtles. Measures require LMP compliance inspection, enforcement, and monitoring of sea turtle orientation. Lighting not compliant with the LMP must be made compliant prior to commencement of the launch/landing/processing operation. The incidental take for CCAFS and for KSC disoriented sea turtles was set to 3% for hatchlings and 3% for adults. The Service concluded this level of incidental take is not likely to result in jeopardy to sea turtle species or result in destruction or adverse modification to critical habitat (USFWS 2017).

KSC and CCAFS continue to make progress in reducing light use through development and implementation of LOMs and LMPs for launch complexes and facilities, and replacement of legacy, short-wavelength lighting with new light-emitting diodes (LED) long-wavelength light fixtures that are less disruptive to sea turtles and other wildlife (L. Phillips/NASA Environmental Management Branch, 2019, pers. comm., and A. Chambers/USAF 45SW, 2019, pers. comm.). A draft LOM specifically for LC-39A was prepared by SpaceX; approval is now pending with the USFWS. The LOM would be updated as necessary to reflect additional lighting and changes to existing lighting operations resulting from construction and operation of Starship/Super Heavy support facilities at LC-39A. The LZ-1 LMP is being updated by SpaceX (A. Chambers/USAF 45SW, 2019, pers. comm.).

Arthropoda do not seem to have received the same degree of scrutiny.

Studies of terrestrial invertebrates have been limited to research aimed at controlling salt marsh mosquitoes, Ochlerotatus taeniorrhynchus and Ochlerotatus sollicitans (Platts et al. 1943, Clements and Rogers 1964). A recent study (2016-2017) of bee population declines in urban environments was conducted with collections from KSC included; however, the report is not yet available (A. Abbate/Auburn University, pers. comm. 2019). A detailed biological survey of terrestrial invertebrates has not been performed on KSC.

Here is a Scott Manley video summarising the report and pointing out details it reveals about the Super Heavy/Starship, its engines, and the planned launch facilities.


Author: John Walker

Founder of, Autodesk, Inc., and Marinchip Systems. Author of The Hacker's Diet. Creator of

8 thoughts on “SpaceX Super Heavy / Starship Environmental Assessment Report”

  1. ctlaw:
    I am surprised about the drone ship. The ship would have to be much larger than the existing ships.

    I didn’t see anything in the report which discussed the rationale for landing on a drone ship as opposed to a return to launch site.  The description of the sonic boom and engine noise footprints for landing Super Heavy don’t appear to be prohibitive—they’re much smaller than the engine noise for launch (which makes sense, since only a few engines are used for landing).  My guess would be that when they worked the numbers, the performance hit of boost back to the launch site outweighed the logistics cost of landing at sea.  On the other hand, this report is based upon the then-current plan for a Super Heavy with 31 Raptor engines.  On 2019-07-21, Elon Musk announced (by tweet, as he is wont to do), that Super Heavy is now planned to have 35 engines.  It is possible this is part of a scaling-up which would provide the extra performance to return to launch site.

  2. Bryan G. Stephens:
    That is a lot of engines.  Seems to increase chance of breakdown.

    That has been conventional wisdom, but Falcon 9 has nine engines on the first stage and has had no trouble, and Falcon Heavy has 27 on the core and boosters, and again no trouble (so far, anyway).  SpaceX always conducts a static test firing of all engines before launch (and says in the report that this practice will continue with Super Heavy/Starship).  That gives a lot of confidence that nothing has gone wrong between the last test firing of the individual engines and their integration on the stage (for example, foreign objects in the fuel tank or propellant lines, problems with valves, etc.).  When the actual launch occurs, all of the engines are lit and the launch commit signal is only sent after all engines are confirmed to be at full thrust and running within design red-lines (chamber pressure, turbopump RPMs, temperatures, etc.).  Historically, most engine problems occur at start-up—once an engine is running stably at full thrust, it’s likely to go on running for its full duration burn.

    The advantage of having lots of engines is that you get engine-out capability.  Falcon 9 can lose one of its nine engines at any time during flight and still complete the mission.  This happened on its very first launch, where one engine had an RUD and the flight continued normally.  Also, smaller engines tend to be more reliable, as they are less prone to combustion instabilities and start transients.  The argument against lots of engines was principally one of weight (since you have duplicate turbomachinery, combustion chambers, and nozzles on each engine) and cost.  But reusability negates the cost argument, since you aren’t throwing away the engines each time.

    Performing a propulsive landing also militates for lots of engines (although not as many as 31).  Real-world rocket engines (unlike those in Kerbal Space Program) cannot be throttled below a certain level, in the case of the Merlin engines on Falcon 9, around 70%.  Since at landing the stage is very light (typically less than 20% of its launch weight), with only a few engines you can’t throttle down far enough to make a smooth landing.  In a normal landing, Falcon 9 only uses one of its 9 engines to land.  I think the plan for Super Heavy is to land on two or three engines, which will provide an engine out capability for landing, which Falcon 9 does not have.

  3. Bryan G. Stephens:
    That is a lot of engines.  Seems to increase chance of breakdown

    Yes, but I think the math goes like this:

    35 rockets with one engine (each) launched 35 times with a failure rate of 1 in 35 will succeed 34 times and be destroyed on the 35th attempt.

    Unlike all 35 of the 35-engine rockets, which all make it to space with one engine (each) on strike.

    Obviously this uses magical statistics, which do not exist.  Really we have no idea what will happen except that after millions of launches, the engine failure rate will be pretty close to 1 in 35.

  4. Bryan G. Stephens:
    I just remember all the N-1 issues.

    The problem with the N-1 is that the Soviets couldn’t afford the facilities or the time to test the first stage on the ground before attempting a test flight.  The NK-15 /  NK-33 engines were reliable by themselves, but there were all kinds of problems integrating that number of engines into the first stage.  Plumbing broke, control circuitry failed, vibration caused other failures, etc.  This caused every flight test to fail during the first stage.  The Saturn V, by comparison, first ran the five engines together in a test stand to verify their interactions, then after testing each engine individually, the engines were installed in the flight stage.  The flight stage was then tested in a full-duration firing before shipping to the Cape.  Basically, if some problem were going to appear, it would have been caught long before the first flight test.

    (There was one exception to this: on the unmanned Apollo 6 Saturn V test flight one of the engines shut down when a liquid hydrogen line feeding the engine broke.  This was due to a phenomenon [air freezing out on the flexible line] which would not happen in ground testing, but only in a vacuum.)


Leave a Reply