Samarth,
I think the basic E & I information you need is in the data sheet. This is an avalanche device, so it must be biased near the breakdown voltage in order to obtain the rated sensitivity. I found the following in the linked data sheet:
Note that a specific value of VR is supplied with each device. When the photodiode is operated at this voltage, the device will meet the electrical characteristic limits shown above. The voltage value will be within the range of 180 to 250 volts.
Maximum Ratings, Absolute-Maximum Values Reverse Current at 22°C:
Average value, continuous operation 200 μA
Peak value (for 1 second duration, non-repetitive) 1 mA
So the power required for the detector diode is very small, perhaps 60mW, or less. If the associated amplifier circuit draws any current from the HV supply, you will need to add that in.
I imagine that the stability and noise requirements for this supply might be very tight.
If you have a budget, and project deadlines, I believe that such inverters can be purchased off the shelf. You may have to settle for a 24V input one and regulate-down your 30V source to feed the inverter. The output may need some additional filtering. Simple L-C or R-C filters may do the job well enough.
If you are just tinkering, you might start with a CCFL inverter, or at least the transformer from one. Or one from BitHead's camera flash. Each disposable camera has on of these.
The power level is so small, that you could even use a boost converter with a simple inductor, rather than a step-up transformer.
For the do it yourself route, I suggest some research on the Linear Technology and Texas Instruments Semiconductor sites. Both make many power supply controller ICs.
Many of the available controller ICs can be used. If the IC has an integrated power switch device, you can add a discrete common base or common gate transistor to get the voltage you need. Or perhaps, Power Integrations makes a part that would work for you.
Ted