In Nature of 18 Februari 2016 at page 263 we read:
"Surprisingly Ligo's first detection dit not come from a binary neutron
star system etc but from two large BH's. Both were of the order of 30
times the mass of the Sun. 'They are real astronomical beast'."

My understanding is that BH's 30 times the size of the Sun are extremely
small.See: https://en.wikipedia.org/wiki/List_o..._massive_stars
See: https://en.wikipedia.org/wiki/List_o...ve_black_holes

At page 262 we read:
"Although the two BH's had probably been orbiting each other for
millions of years LIGO began to pick up their waves only when they
reached a a freq of 35 Hz. This frequency rapidly increased to 250 Hz."

This sentence gives the impression that binary systems in general
are stable configurations. So what happened.
Simulations I have performed give the impression that for a binary
system to merge generally speaking at least one object should increase
in mass. For example comets which collide with the Sun will decrease
the size of the solar system.
This brings me to my question: Are we sure that this is really
a binary system and is not a third object involved.
See for example:
http://users.telenet.be/nicvroom/VB%...0operation.htm
Starting point of simulation is a binary star system.
The standard configuration is that the masses are 1000, 1000 and 1.
When you change this configuration to 1000, 500 and 50 you can
investigate the influence the a relative large third object.
In such a simulation you can get a configuration with average
distance (radius) of 200 units, which changes in an elipse which
shortest distance of 100 units.

In the case of the BH you will get something like 36,29 and 3 Solar m.

In this particular case I assume that the third object is rather small
but it also could be a red giant which constantly transmit mass
to the binary BH system

Nicolaas Vroom